Speed Limit Fact Sheet ( PDF )
Free to download, print & distribute.

By Charles E. Dornsife January 9, 2000

This was a paper that was submitted to the Transportation Research Board Annual Meeting in 2000.

1966 Law, Best Guide for 21st Century
Re: Safety and Setting Speed Limits


In 1966, Congress passed a law that required all traffic control devices on public roadways in the nation be based on sound engineering principles, practices and have a common “basis in fact” determination, appearance and application.

These mandates are found in Title 23 , United States Code, Section 109(d) and Title 23, Code of Federal Regulations (CFR), Part 655.601 through 655.603 (Update: Title 23 C.F.R. Part 655 Subpart F- Traffic Control Devices on Frderal Aid and Other Streets). The MUTCD (Manual on Uniform Traffic Control Devices) is where these requirements are specified. The goal was to return safety decisions to those that base their decisions only on findings of fact that have proven to make us safer; with uniform nationwide continuity of meaning and expectation.

They expanded the FHWA domain over construction of the nation’s highways to include all signs placed on them. Entrusting this responsibility solely to the licensed traffic engineering professionals and their institutions, because the tenets of their profession require studies to test a thesis, peer review and verification before a standard, practice, procedure or principle can be incorporated into their engineering body of working knowledge.

This law outlawed the old practice of each political entity in the nation, based on local tradition or conjecture, deciding the appearance of “traffic control devices” and what, where and how such devices would be used in their town, county or state.

Congress granted itself special powers in response to a national crisis not otherwise contained in the law when they adopted the NMSL (National Maximum Speed Limit). When the federally mandated NMSL was repealed, state politicians wrongly assumed that this failed policy of setting arbitrary limits was a political issue deferred to them.

This simply isn’t the case; the repeal returned the nation to prior law. This law has matured and authorization for a locality or state agency to use a federally regulated “official traffic control device” is contingent upon full compliance with the Code of Federal Regulations and its MUTCD. It no longer allows the use of regulatory traffic control devices to post politically invented numbers.

The national MUTCD section 2B-10 (update: 2B-13), “Speed Limit Sign” federal designation R2-1, is a regulatory safety device; therefore, the numeric value shall be determined on basis of an engineering study finding that determined a need exist for that particular section of highway.

The NMSL legacy of numeric values being determined by political decree has understandably left States in confusion with who has authority over traffic engineering issues. Compounded by the myth our government created to support the NMSL that equates lower speed limits with safety; this has all but institutionalized unreasonably low limits and with incorrectly applied or absent traffic control devices that directly contributing to increased accident rates while making violators out of safe drivers.

We need to return to the principle embraced in this well thought out 1966 law that is still in full effect, empower the engineers to do what is right and again remove this issue from the political arena. This law to Motorists is a “Bill of Rights” that should not be altered. It is our best hope to meet the safety needs of the state and it also protects citizens from unwarranted prosecution. In support of this endeavor this paper should contribute to a clearer understanding of how speed limits are supposed to be set, including factors to be considered and their impact on safety.


This is a classic case of the unintended consequences of good intentions, and the nature of government.

In 1966, Congress passed a momentous law that required public safety issues to be based on findings of fact and uniformly applied across the land. The law created the NHTSA (National Highway Traffic Safety Administration) with safety on our nation’s highways as its prime directive.

Eight short years later, Congress turned this agency into a propaganda arm of the government where they “knew what was best for the people” despite the fact that their good intentions were contrary to human nature, advancing technology and the findings of the safety engineers.

After 25 years of enormous cost to the nation, NHTSA created an Urban Myth that equated lower speed limits with safety. The evidence at every stage of the process was indicating neither speed limits, nor that enforcement had any meaningful effect on traffic speeds and accident rates.

Speeds have been going up since the automobile was invented in the 19th century. Accident and fatality rates have been going down and they have continued to do so since the beginning of the last century. The public after 25 years accepts the basis for the myth; but assume it is targeted at the other crazies because when they drive over the posted limit, they are driving safely – which they are.

This process has conditioned motorists to disregard speed limit signs at wholesale levels. In the name of safety, Government agencies have also been allowed to unwisely use stop signs, double yellow lines etc. Again, because the public has found that violating them represents a low risk for being involved in an accident, they are now disregarding these devices too. The people being regulated do think; they are not sheep, when they realize there is no justification; they act accordingly. Like crying wolf, this has lessened the effectiveness of all traffic devices where there is a real need and justification.

This speed myth that NHTSA created, has become so pervasive that traffic engineers trying to do what they know is best (raising limits to reasonable levels), are vilified and attacked by politicians, the press and citizen groups demanding lower limits because they wrongly believe they are safer.

It is time to accept the fact that increases in speed are the natural byproduct of advancing technology. Therefore, the focus needs to be on flow management and making sure the devices are only used when they have a real expectation of accomplishing their desired effect.

If the object of this TRB panel is to restore the credibility of our nations regulatory traffic control devices, the committee might concur to endorse the MUTCD and recommended that the engineering profession be allowed to fulfill the promise of the 1966 Congress’s exceptional Traffic Safety Law.


This section is a synthesis of the most current engineering positions on each of the issues to the best of the author’s knowledge.

Basic Tenets of Speed Laws

Laws protect the public by regulating unreasonable or unsafe actions. Actions of a reasonable person should be legal. Most people drive in a safe and reasonable manner. Laws cannot be effectively enforced without the public consent and voluntary compliance. Give highway agency or localities authority to post limits (higher or lower) on the basis of engineering study and if set to indicate the maximum reasonable and safe speed. And that they be fair, be related to risk, be credible and accepted by drivers and enforce control over unreasonable behavior.

Speed Limit

Numeric value that speeds-in-excess-of represents an unacceptable hazard to others. Specific procedures must be met to determine this value, utilizing site specific “Basis in Fact” findings.

All Speed Limits Should Be Prima Facie*

Prima facie limits recognize that conditions vary during a day and subject the motorists to penalties only when their speed in excess of the limit is a hazard to others under the conditions then existing with regard to the actual and potential hazards. Prima facie limits also acknowledge that even when limits are set at reasonable levels, there is a band of speed greater than the limit that is safe and no one speed is safe for all conditions.

*Absolute limits are warranted in school, construction and hospital zones, business districts and alleys; high pedestrian or vulnerable activity areas. However, absolute and special speed zones shall only be in effect during specific times of day when the hazard is present and removed immediately when the hazard ceases to exist.

Speed Limit Sign

A federally regulated “Regulatory Traffic Control Device” (R2) that local and state authorities are granted a “conditional use permit” to use as long as they comply with the mandates contained in the National MUTCD (Manual on Uniform Traffic Control Devices).

MUTCD section 2B-10 (update: 2B-13) requires that the numeric value be determined on basis of an engineering study finding that determined a need existed for that particular section of highway. To wit, review of roadway’s prevailing speeds, highway conditions, hazards and accident history – determining if the use of this mitigation device is justified, if so, then report on what numeric values are authorized within the guidelines to post. Once determined, the political entity can choose between those options that have been identified appropriate for that particular section of roadway.

There is no engineering basis that requires speed limits. In fact, the opposite is true. The MUTCD admonishes that devices only be used when warranted and when they would have the desired effect on reducing accidents. The MUTCD states, “A standard device used where it is not appropriate is as objectionable as a nonstandard device; in fact, this may be worse, in that such misuse may result in disrespect at those locations where the device is needed.” The MUTCD also recognizes urban and rural conditions may require different determination criteria.

Engineering Study Required To Post Speed Limit Sign

Periodic traffic engineering studies are required on all roadways – even on roadways without speed limits. The measurement of speed in a traffic engineering study is only one component of an extremely important periodic safety evaluation of a particular section of highway. When this information is combined with the accident history for that section of highway the engineer has the information needed to determine what mitigation efforts may be required to address the identified safety hazards and flow management issues and needs for that section of roadway.

The measured prevailing speed of traffic determines the size and the distances for all traffic control devices and where they need to be placed to accomplish intended desired action, the timing of traffic lights, the length and capacity of turn lanes, the type of signs that may be warranted to guide traffic or warn traffic in advance of hazards that aren’t readily apparent to the approaching motorists et al.

MUTCD: 1A-4 Engineering Study Required

The decision to use a particular device at a particular location should be made on the basis of an engineering study of the location.

2B-10 (update: 2B-13) Speed Limit Sign (R2-1)

The Speed Limit sign shall display the limit established by law, or by regulation, after an engineering and traffic investigation has been made in accordance with established traffic engineering practices. The speed limits shown shall be in multiples of 5 miles per hour.

In order to determine the proper numerical value for a speed zone on the basis of an engineering and traffic investigation the following factors should be considered:

1. Road surface characteristics, shoulder condition, grade, alignment and sight distance. 2. The 85-percentile speed and pace speed. 3. Roadside development and culture, and roadside friction. 4. Safe speed for curves or hazardous locations within the zone. 5. Parking practices and pedestrian activity. 6. Reported accident experience for a recent 12-month period.

Accepted practices: “Established traffic engineering practices” are those practices that are nationally recognized and have a substantiated basis in fact.

Nationally recognized: The engineer must be able to articulate what studies or FHWA, ITE or ASSHTO recognized guidelines, standards or studies that this basis-in-fact finding is based on. Interstates and surface highways each have their own different studies, standards and practices.

The authority ordering the placement of the regulated devices must be applying nationally recognized practices. If there is no staff personnel that understand these mandates and procedures, then the MUTCD advises them to consult with outside experts or engineers from other agencies.

The use of the term “Engineering Judgement” is not a license authorizing unapproved or hypothetical unverified methods, unfounded conjecture or to carry out personal or collective beliefs. There is a forum to ask for a variance from nationally accepted practices if they feel they have a better answer or method, but the National Committee prior to field trials must approve this thesis.

A “Basis in Fact” finding can not be made without a site specific review/study that includes a detailed review of the above factors that are applicable for the particular section of roadway being regulated. Items to be reviewed are different for different classes and types of urban and rural highways, but site specific studies are required to justify the use first, then what particular devices are indicated and where the would be most effective and then the numeric value assigned must be supported and germane to accident reduction at that location.

Speed Limits Are Not To Be Lowered For Conditions Readily Apparent Or To Be Used As Curve And Hazard Warning Devices

Posting for curves or hidden hazards is to be done with advisory warning devices and downward posting for those conditions already apparent is not advised. Speed limits are posted for the free-flowing higher prevailing speed sections. Curves and hazards are to be posted with advisory warnings in proportion to the danger they represent to the approaching motorists travelling at the prevailing speeds – at sufficient distance to timely advise the motorists of the degree and nature of the approaching hazards.

California Vehicle Code states the anti-downward zoning rationale well:

“It is the intent of the Legislature that physical conditions such as width, curvature, grade and surface conditions, or any other condition readily apparent to a driver, in the absence of other factors, would not require special downward speed zoning, as the basic rule of section 22350 is sufficient regulation as to such conditions.”

Cost Of Compliance To These Regulations Requiring Engineering Studies

These practices are already being certified as being done in exchange for federal highway funds received by the state. Nevertheless, many highway agencies and localities claim it would be too great a burden to actually comply with the law.

It is not a burden on the agencies. It is rather a choice made in the allocation of resources by the posting authority. For a proper study to be done on a cross-town expressway. As an example, it may take several man hour weeks once every 5 years. On the other hand, there has been no shortage of man hours available for enforcement with multiple officers assigned to work an area, frequently on a daily basis, resulting in many man hour years being allocated over this same 5 year period for enforcement of limits not lawfully set.

By far the most effective use of funds have been the investments made in engineering reviews along with mitigation efforts to guide and facilitate traffic by reducing flow conflicts and removing hazards. This is the primary function of an engineering study and the data collected to verify the appropriate speed limit posting is only one component of this necessary safety review program.

Free-Flowing Speed

Free-flowing is a minimum 4 second unencumbered headway between vehicles. Taken at representative higher speed free-flowing sections of the roadway being studied. Speed measurement locations should not be encumbered or under the influence of intersections, curves (horizontal or vertical), lane reduction zones, hazard warnings, traffic signals or stop signs.. i.e. choke points in measured speeds or flow conflict points. Minimum 500′ distance is required from a flow obstruction to the location where the speeds are measured.

The lead vehicle in a queue skews down results by unduly influencing the following vehicles natural pace speeds. Therefore, all vehicles in queues are not to be counted because the lead vehicle has biased down speeds all those that follow. If you count the lead vehicle and not the others in the queue you bias down the speed proportional by the number of vehicles who’s higher speeds would have been measured if the slower vehicle had not impeded their movement.

The intent of the speed measurements is to determine the actual speed of the unimpeded traffic. The speed of traffic should not be altered by concentrated law enforcement, or other means, just prior to, or while taking the speed measurements.

85/67th Percentile Speed

Defined as that speed at or below which 85 percent of the traffic is moving. On urban roadways the 85th percentile speed has been found to be the safest speed, where the 85th exceeds 50 mph the safest speed shifts to the 90th percentile. Speed limits established on the basis of the 85th percentile on urban roadways conform to the consensus of those who drive highways as to what speed is reasonable and prudent, and are not dependent on the judgement of one or a few individuals.

However, in no case should the speed limit be set below the 67th percentile speed of free flowing vehicles.

Determining The Safe For Conditions Speed

The traffic engineer’s speed survey is simply a measurement of the “public’s consensus” as to what free-flowing speed they have found to be safe. Each motorist drives at a speed they feel comfortable and safe with. Each driver expresses their comfort level by their actions, taking in to account all visual clues that may be present. Engineers have found this to be a better process than basing the speed limits on the arbitrary judgments of a few. Setting all traffic control devices, turn lane lengths etc. based on this measured public consensus has been the most effective in reducing accident rates.

Back to Table of Contents


This section is a synthesis of the most current engineering positions on each of the issues to the best of the author’s knowledge.

The overwhelming evidence supports the 85th percentile speed as being the best indicator of what speeds have been found appropriate by the highway’s users on these classifications of roadways. It is easily obtained, applicable to all roadways in this classification where the 70-90th percentile speeds have been found to be the safest – lowest point on the “relative accident involvement risk curve”.

Surface streets, arterials and expressways:

85th rounded up to the next 5 mph increment. (FHWA workshop)

Residential streets:

85th rounded to the closest 5 mph increment. (FHWA workshop)

Setting speed limits and all traffic control devices to guide vehicles traveling at the 85th percentile of the prevailing free-flowing speed measured during optimum conditions;

Results in limits found to be acceptable by the majority of the motoring public, and

Has found to lower risk and accident rates, and

Doesn’t significantly change the speed of traffic, and

Highest speeds do not change or may be lowered, and

The highest risk slow drivers increase speeds reducing the number over takings, and

Maximizes the carrying capacity of a roadway, thereby reducing congestion and associated pollution.

Guidelines For Determining 85th Percentile Speeds

Prima facie limits: (FHWA workshop)

85th percentile determined by 24 hour composite of free-flowing traffic.

Absolute limits: (FHWA workshop)

85th percentile determined by highest free-flowing speeds in a 2 hour bin during a 24 hour period. Speeds do vary from peak and off-peak periods by 4-8 mph over a 24 hour period.

Some jurisdictions do allow counts of 100 free-flowing vehicles and make further exceptions allowing 50 vehicles to suffice. Allowing 100 vehicles counts on roadways carrying several thousand vehicles an hour is not sound practice under any standard, particularly when this state’s law accepts this as certification to set the limits unbelievably for up to the next 10 years. With modern equipment there is no reason to accept such small samples and this is why professional guidelines must be adhered to; ITE says no more than 5 years between studies and the FHWA sufficient sample size recommendations.

Secondary justification to lower below the 85th percentile:


The ITE Speed Limit Sub-Committee 4M-25 has a wider acceptable range for surface streets; nevertheless, found that regardless of secondary conditions that in no case is a posting under the 67th percentile acceptable.

Methods of measurement:


Loops, dual rubber hoses, imbedded magnetometers and devices like the Nu-Metrics Trans Star 9700 which is a small computer that is temporally anchored to the pavement. The software for these programs can produce vehicle speeds, classifications and headway.

Radar and Laser Not Advised:

As long ago as the late 70’s the FHWA found that states using radar to measure speeds for compliance with the National Speed Limit were showing measurably better compliance than those states using traffic loops. They outlawed the use of radar to collect data in compliance reports.

The placement of a vehicle on or near a highway using radar has shown to bias down traffic speeds. Those without radar detectors do slow in response to an unknown vehicle parked on a shoulder and vehicles with detectors adjust their speeds lower up to a mile or more prior to the radar/laser transmitters. This has been shown to bias down studies up to 6 mph. Another problem with radar/laser is the cosign angle error factor, as you increase the angle to the approaching vehicle there is an exponential lowering of the measured speed. Angles greater than 15 degrees should not be used.

These methods are subjective, put the operator at risk and because they require an operator many locations suitable to measure from are not available on many roadway sections being surveyed. Operators tend to go where it easiest to measure from, which many times are the very flow conflict points they are admonished to take their measurements clear of.

Another caution, to obtain direct measurements with low cosign angles, finding sections on mountain roads long enough to make sure that your measurements points are at least 500′ from the influence of a curve or flow restriction point becomes more difficult.

Accident Reports:

The study shall include a report on the number and nature of accident for that particular section of roadway. The engineer should compare this with an on site analysis of the location looking at all contributing roadway design, delineation and traffic control device factors present.

Frequency of studies:

Statutory time: ITE guidelines state highways should have safety reviews at least once every 5 years and this includes an engineering study of current traffic patterns and accident data.

Exception: Anytime either development or highway work has been done in area. Simply resurfacing a highway without other changes can result in increased measured 85th percentile speeds up to 4-6 mph.

Number of speed monitoring locations and speed zone changes:

You want to minimize speed limit changes by averaging the findings on contiguous sections of highway to post uniform speed limits. If the posted limit does not exceed a plus or minus 5mph posting recommendation for this section, averaging is warranted. If the speeds show an incremental increase in a particular direction then post the limits accordingly.

The FHWA workshop recommended measurements be taken every 1/2 mile for local roadways. There should be sufficient measurement points to get a true picture of the prevailing speeds.


This section is a synthesis of the most current engineering positions on each of the issues to the best of the author’s knowledge. There has only been a few comprehensive studies of urban interstates, Cirillo’s speed relative risk curve became the accepted baseline. (Cirillo, J.A., Interstate System Accident Research Study II, Interim Report II, Public Roads, Vol. 35, No. 3, August 1968, pp. 71-75) The recommendation to use the Indiana study was part of an FHWA workshop handout. (Maximum Speed Limits: A Study for the Selection of Maximum Speed Limits, Indiana University, Institute of Research in Public Safety, final report, volume I, Oct. 1970, Contract No. FH-11-7275) Opinions: US policy and NMSL comments, Variable limits, Cirillo curve, Personal observations and California data collection over a several month period.

If speed limits are found to be warranted for urban Interstates, it must be noted that when the Cirillo relative risk curve is overlaid on actual speed data, the safe for conditions speed at these location extends up to and includes the 95th percentile and more.

Therefore, at a minimum, the limits should be Prima Facie and set at the 1970 Indiana study recommendation of the 85th rounded up to the next 5-mph increment, until such time as this subject can be further studied.

Variable speed posting should be used primarily as a flow management tool displaying optimum recommended pace speeds to maintain flow and guide traffic; rather than (ineffective) regulatory speed limits during congested or heavy traffic conditions.

Truck and passenger vehicle limits need to be set independently. That being said, the closer the truck limit is to the passenger vehicle limit the better. Trucks should be restricted to the right 2 lanes except at highway convergence areas.

What We Do Know About Limited Access Interstates And Freeways

The legacy of the NMSL left us with few government personnel aware of prior law or understanding of its principles and mandates or knowledge of how to set interstate limits. Information on procedures is sparse and there are many misconceptions prevalent, even among engineers.

There is a direct correlation between accident rates and flow conflict points; converging highways and interchanges. The accident rates at these locations are primarily a design question regarding methods used to allow traffic to egress and ingress into the stream or converge. If they are too close to each other rates increase exponentially. Once the stream clears these conflict points the accident rates drop precipitously.

The speed of traffic is self-regulating, flow becomes uniform as congestion increases and the speed of traffic in not effected or influenced by posted speed limits nor are the nature and type of accidents that do occur.

Conversely, when the conditions become light, you do see increases in speed differentials and during these light free-flowing periods’ accidents for any reason drop to negligible rates.

The shape of the Cirillo curve was heavily influenced by the accident rates of vehicles that had slowed or were traveling slower than the mean, in the lanes closest to flow conflict areas. That being said, this report found that vehicles traveling at the mean plus 12 mph were the safest and until you exceeded the mean plus 18 mph did the higher speed vehicles reach the risk of those traveling at the mean or slower.

When site data is examined and only those accidents during the light free-flowing periods away from the conflict points are analyzed, flow differential accidents are extremely rare events. The lower-than-US fatality rates on the Autobahn (where flow management is the primary safety strategy) would indicate that having speed as the cornerstone of the US policy is misplaced.

In Southern California, Caltrans has placed flow monitoring stations on all their major freeways. At many of these locations, the data is available to the public live via the Internet. In addition to reporting the average speed, in the past they you could get the details of this aggregate in a separate live attached report. All the areas in the report are posted with 65-mph speed limits.

The average speed during light traffic conditions at some locations consistently ranges 10-15 miles over the posted limit, and at times 20 plus mph over with volumes up to and greater than 6,000 vehicles per hour. The detailed reports routinely registered number 1 and 2 lanes average speeds up to and over 90 mph day after day with no corresponding accidents. These same conditions have been observed at many other locations within California.

The question: If the speed limits are heavily enforced there, and the motorists continue by their actions and low accident rates find speeds up to 20 mph and more over the limit to be safe. What should the speed limits be set at? To what end?” If speed caused accidents during light free-flowing conditions are rare occurrences.


There are no studies on contemporary Rural Interstates and Primary highways for anyone to use as a baseline or to refer to make a judgment from. The author’s comments are from personal contact and ongoing communication with the Montana Department of Transportation during the entire period shown in the charts and these discussion are ongoing today. A special note of thanks to Jack Williams, Montana DOT. His assistance in sorting through this data has been invaluable.

Montana’s DOT undertook many different evaluations to determine if there were cause and effect relationships to accidents. They studied the origin of the vehicle registration and speed to see if there was a correlating accident cause and to determine which group had the highest speed. This honor went to vehicles registered in Alberta and Washington State; they were not over represented in the accident data. The following is the conclusion of the author based on what evidence was made available and the accident data supports these observations.

The 85th percentile is not a sound basis to determine these limits. The evidence provided by Montana’s experience would suggest that speed limits have no effect on fatality or incapacitating injury rates for this class of roadway.

If speed limits are to be set then they should be set to encompass all those vehicles found on the relative risk curve to be a low risk for being in an accident.

Applying numbers to the most accredited relative risk bell curve, by Cirillo, tells an interesting story. This curve was based on urban interstate research during off peak hours. It found the safest speed to be the mean plus 12 mph, and when charted on the graph the safest speeds extended up to the mean plus 18-20 mph.

Translated, the current mean on most rural interstates is 75 mph and the safest speed would be 87 and the vehicles least likely to be involved in an accident are those traveling between 75 and 93-95 mph.

The current speed limits are set so low that the vehicles least likely to be involved in an accident, according to the studies, are the primary targets of enforcement during the best highway conditions.

Should the limit be set at 90 or 95 mph? Either way, it would seem to have no effect on accident rates. The Montana and Autobahn real life experience of unposted limits having no apparent effect on safety can not be ignored.

What We Do Know About Rural Interstates And Highways

After a 20 plus year period, with unjustified low speed limits, heavy enforcement and supporting government propaganda Congress repealed this failed law. Shortly thereafter, one state provided traffic engineers with an extraordinary opportunity to test their long held findings, that posted speed limits will not significantly change traffic speeds or accident rates with one notable exception. Lowering limits (below 85th/67th) has shown to have an inverse effect on safety, contrary to popular belief.

After the repeal of the NMSL, Montana went to NO SPEED LIMITS outside of urban areas ON ALL ROADWAYS. Highway fatality and incapacitating injury rates remained within normal deviations. More interestingly to the discussion of speed limits, the “basic speed law” was ruled unconstitutional and the public knew they had no enforceable speed limits whatsoever. Fatalities were down during this period in 1998-99. The fatalities in 1999 were the lowest yet for the first half of the year, during the second half it has gone up. It appears at this point that fatalities 1999 will be lower than those for 1998.

Montana FARS Report (Fatal Accident Reporting System)

  Rural Interstate Rural Primary Total of Both % State Fatal Total Statewide
1994 41 51 92 50.5% 182
1995 33 48 81 34.5% 186
1996 39 54 83 46.4% 179
1997 51 56 107 48.0% 223
1998 30 54 84 40.3% 208
5 yr. mean 39 53 89 43.9% 196
(Fall 95 – mid 99 no daytime speed limits)

Even though there was a dramatic drop from 1997 to 1998 with the lowest number of fatal accidents reported in modern times on Montana’s interstates, the 1995 number shows this is within normal deviation. Therefore there is only one conclusion that can be drawn from Montana’s real life example, there is no apparent correlation between the number of fatalities, enforcement and posted limits on rural roadways.

The 1994 data was below the average for the previous decade. The ’94 percentage of fatalities on rural interstates and primary state highways was greater in relation to the state’s totals than the no speed limit period.

Caution these are static numbers. Fatality rates are based on vehicle miles traveled, the nations traffic volumes are up and higher risk personal discretionary travel is at record levels. The database of Montana’s relative VMTs appears to have significant inconsistencies yet to be quantified. Therefore this important component is missing, making firm assumptions unwise, except that the traffic volumes have been increasing at some undefined rate.

One of the primary functions of Speed Limits is to decrease the number of overtaking, thereby reducing accident rates. The author asked Montana DOT for data on multiple vehicle accident to see if there were any changes or trends.

Montana: percentage of daytime accidents involving multiple vehicles

Rural InterstateS Rural Primary
1994 24 % 53 %
1995 26 % 53 %
1996 29 % 52 %
1997 25 % 50 %
1998 22 % 49 %
5 yr. mean 25 % 51 %

(Fall 95 – mid 99 no daytime speed limits)

On this classification of highway, this thesis of a change in accident frequency would not be supported by the Montana’s experience.

The information was asked for because the author’s personal observations indicated that slower traffic was yielding the left lane by keeping right and/or moving closer to the shoulder to allow safe overtaking. Instead of increasing accidents, with the expectation of higher speeds, it would result in better lane courtesy. The expectation of higher speed phenomenon has also spilled over to seat belt use. They are now reporting a 88% compliance rate on their interstates and primary state highways, with only a secondary enforcement law. Actual measured speeds are equal or lower than monitoring station data collected from agencies in other western states.

In examining the data from Montana there were areas that made no sense whatsoever, until it was explained that in 1996, the Highway Patrol changed the accident reporting form and procedures. The number of reported accidents statewide went up significantly after that, but there were no corresponding increases in fatal or incapacitating injury accidents.

The Chief of the patrol was campaigning endlessly to get speed limits and the most probable answer is he instructed his officers to make out reports on all incidents, where in the past they may have been less demanding. Therefore, the only data used in this report are the fatality accidents, because of their nature, they have no political component.

Weather variations from year to year in Montana can be significant and weather does influence the number accidents and type. Three years after its return to no daytime speed limits, they ended with the lowest percentage of multiple vehicle accidents, fatalities and incapacitating injuries recorded in modern times on its interstates and respectably low numbers on its rural primaries.

Montana’s conditions are common to other states. During this 5 year period, San Diego’s 65 mph posted urban interstates were consistently reporting higher speeds (with negligible accident rates) than Montana’s rural interstates (no speed limit), at volumes per hour that would take Montana days to achieve. These higher than Montana flow speeds were also observed at other urban interstate locations during this period.

The Montana experience solidifies the long held engineering axiom taken directly from the Washington State DOT website: “that people don’t automatically drive faster when the speed limit is raised. These are common misconceptions, along with the mistaken belief that speed limit signs will decrease the accident rate and increase safety, and highways with posted speed limits are safer than unposted highways”.


All speed limits should be prima facie. On urban surface streets, highways and expressways the evidence supports 85th rounded up to the next 5 mph increment and on residential streets should be rounded to the closest 5 mph increment. No other considerations should be considered on roadways in this category to lower the limit. On surface streets there may be other mitigating factors but in no case on these roadways should it be set below the 67th percentile. Once the prevailing speed is determined, regardless of the final posted numeric value, all traffic control device applications should be set to manage traffic traveling at the prevailing speed (i.e. sign size, signal timing. turn lane lengths, pre-marking of intersections).

Interstate speed limits should encompass all those found to be driving within the safe for conditions speed as determined by the Cirillo relative risk curve. Because of the nature of traffic speeds, when compared to densities and corresponding accident rates during these variable periods, speed limits are providing little safety benefit. At a minimum (until this can be further studied), limits should be set at the 85th rounded up for that particular section of highway and vary accordingly in relation to the studies findings for each particular section.

On rural interstates and highways the evidence shows that fatality and incapacitating injury rates didn’t changed when all limits were removed. If we set limits to encompass all those that have been found by the public’s consensus to be driving safe for conditions, do we want to post limits at 85, 90 or higher? Traffic speeds did not significantly change, people in these regions continued to drive at speeds they were comfortable with, which interestingly enough is at speeds lower than their counter parts on high density urban freeways with low posted limits.

These highways, on a one-by-one basis, should be returned to no limits. If the experience remains the same, the program should be expanded. At the same time, the state should be focusing on infrastructure improvements that manage flow and hazard removal, while educating the public on slower traffic keeping right and highway courtesy. Human nature in Montana brought seat belt usage on its rural interstates and primaries up to approx. 90%, slower traffic has been yielding the left lane and highway courtesy has increased.

As a believer in scientific methods, as was the 1966 Congress, we as a Nation would be best served by entrusting this responsibility solely to licensed traffic engineering professionals and their institutions. The tenets of their profession require studies to test a thesis, peer review and verification before a standard, practice, procedure or principle can be incorporated into their engineering body of working knowledge.

The law is already in place. It works when it is applied and it meets the objective of the TRB workshop. Make recommendations that will further empower and aid it.