California Safe Speeds Toolkit: Research on Speeds, Speed Limits and Safety

1. What is Safe Speed? 

Speed is a significant concern both locally and statewide. According to Caltrans, more than 3,600 traffic fatalities and 13,000 serious injuries occur in California each year (Caltrans 2022). Approximately 34% of these fatalities and serious injuries are related to speeding and aggressive driving (Caltrans 2023).

A safe speed is a travel speed that lowers the likelihood of a crash by allowing more response time and better visibility. It also accommodates human mistakes and provides crucial redundancy by maintaining impact energy on the human body at a tolerable level in case of a crash (FHWA 2022).  The Federal Highway Administration (FHWA) states that Safe Speeds reduce system kinetic energy and accommodate human injury tolerance (FHWA 2022).

In opposition to the traditional view that drivers can choose reasonable and safe speeds (see Section 2), safe speed limits should instead be set based on “the likely crash types, the resulting impact forces, and the human body’s ability to withstand these forces” (NTSB 2017). This approach minimizes the likelihood of fatal and serious injuries for all road users, particularly vulnerable road users such as pedestrians and bicyclists (NTSB 2018; Forbes et al. 2012; Jurewicz et al. 2014). For many jurisdictions across California, this often involves reducing the posted speed limit to a safe speed limit, which is a proven safety countermeasure according to FHWA (FHWA 2022) (see Section 3).

Determining safe speed limits requires considering numerous factors, including potential hazards, the road environment, the presence and movement of different road users, surrounding land uses, and policy goals for traffic safety. Additionally, a safe speed limit cannot be determined in isolation of road design, vehicle design, and the anticipated road users – all elements of a holistic safe system approach. Local jurisdictions can specifically integrate safe speeds and safe road design (see Section 5) to comprehensively improve safety for road users of all modes, ages and abilities. Without the ability to employ this more holistic approach, speed limits can sometimes be set higher than is considered safe. For example, California speed limit setting procedures pushed Los Angeles to primarily rely on 85th percentile speed data when updating speed limits citywide in 2017, which actually led to speed limit increases on 90 miles of roadways, including several with histories of serious or fatal crashes (Toda 2018). In contrast, Portland and Seattle both adopted more flexible methodologies that incorporated factors such as crash statistics and road type related to the streets under consideration, and speed limit increases were not an observed outcome in these non-California cities (Toda 2018).  

In response to issues such as those experienced by Los Angeles in 2017, laws and guidelines regarding speed limit setting at the state level have been evolving. Recent state legislation has given local jurisdictions additional flexibility to consider factors other than the 85th percentile when setting speed limits. The "Current Speed-Limit-Setting Law" page provides information on various options for local jurisdictions to flexibly set safe speed limits without relying exclusively on the “85th percentile rule.” 

Finally, this toolkit describes several cases where safe speed limits have been systematically implemented. The "Real-World Examples and Self Quiz" page provides relevant examples of safe speed limit implementation in California.

2. Limitations of the 85th Percentile Rule for Setting Speed Limits

The “85th percentile rule” was developed in the 1930s and became a widely accepted “conventional wisdom” for setting speed limits across the United States. This approach sets the speed limit based on the actual speed of drivers at the 85th percentile (that is, the speed that 85% of drivers do not exceed); it assumes that the collective judgment of drivers determines what is safe. Proponents have historically argued that it reduces the variability in travel speeds and allows law enforcement to focus on the most extreme cases of speeding (Grembek et al. 2020, 27).

An underlying assumption of the “85th percentile rule” is that drivers can and should set safe speed limits. However, recent research, such as Section 4.1 of the UC ITS report by Grembek et al. (2020), increasingly sheds light on the behavioral challenges for drivers (Grembek et al. 2020, 46). Streets in urban and some suburban areas present complex challenges for drivers to self-regulate safe speeds; NCHRP 17-76 concluded that the 85th percentile speed is not the safest approach under many urban conditions (Fitzpatrick et al. 2021). Importantly, urban and some suburban streets may need to accommodate not just drivers, but also pedestrians, bicyclists, and other street users. As such, the traditional "85th percentile rule" is not always the best approach for setting speed limits that align with Complete Streets, Vision Zero, Safe System, Active Transportation, Sustainable Transportation, and Transportation Equity policy goals of local, regional, and state jurisdictions. 

  • Behavioral research shows that drivers cannot self-regulate a safe speed easily on lower-speed roads. Research has shown that drivers can underestimate speed by up to 30% at 35 mph (Grembek et al. 2020, 46). These lower-speed roads often lack strong visual cues such as guardrails and shoulder widths for drivers to assess safety and speed (Ben-Bassat & Shinar 2011). When drivers believe that speeding does not threaten safety, they have a tendency to exceed the speed limit (Mannering 2009).

  • Drivers’ underestimation of speed worsens in adverse weather conditions. Research has shown that the driver’s perception of speed decreases in fog. As a result, drivers think they are driving far more slowly than they actually are in foggy conditions (Snowden, Stimpson, Ruddle 1998).

  • Many drivers speed and believe that excess speed does not threaten safety. About four in ten respondents to a recent survey conducted by the American Automobile Association (AAA) admitted to driving ten miles per hour or more above the speed limit on residential streets, even as 90 percent of them reported being somewhat or completely disapproving of that behavior (AAA Foundation for Traffic Safety 2019).

  • Speed enforcement in local areas is limited and does not create effective deterrence. Speed enforcement in local areas is typically more limited than on highways. Accordingly, the perception of the certainty of punishment is reduced and the overall impact of legal sanctions as a deterrence for speeding on local streets is diminished. 

  • Local area streets are impacted by spatial speed creep from neighboring highways. Research has shown that higher speeds on some highways can cause higher speeds on connecting local area roads (Casey & Lund 1992) and also that pedestrian fatality rates are elevated on roads adjacent to highways (Nehiba & Tyndall 2023). The implication of these studies is that the impact of speed limits on highways can be carried over to local area streets and should be considered as a safety issue.

  • The safety risks associated with speed are higher for pedestrians and bicyclists.The National Transportation Safety Board (NTSB) reports that mitigating speed is vital to improving pedestrian safety (NTSB 2018), and there are unique safety considerations when considering biking as well. Those traveling using active transportation modes such as biking and walking are at great risk of serious injury in the event of a crash with a motor vehicle given vehicle size and speed relative to the bicyclist or pedestrian. Safer speeds can not only reduce the risks to those walking and biking on the streets but also to the operators of motor vehicles who are involved in a crash given the strong relationship between speed and injury severity.

  • The 85th Percentile Rule may lead to speed creep in specific situations. Consider a situation where, collectively, drivers elect speeds such that about half of them drive faster than the speed limit.  This behavior, if coupled with a periodic application of the 85th percentile rule, may cause an upward drift in speeds (Grembek et al. 2020, 44). This sequence of events leads to what is known as “speed creep.” It may be important to systematically distinguish roads that are experiencing speeds that are higher than what is safe from other roads where speed limits are set accurately to better identify which roads are in need of more flexible methodologies for setting speeds. It is also important to note that just a one mph increase in the observed 85th percentile speed could lead to an increase in a speed limit by 5 mph if the one mph increment changes how the speed 85th percentile speed is rounded. Consequently, this could also lead to speed creep.

3. State-of-the-art Evidence on Speed & Safety

The connection between speed and safety is well-established from the perspectives of both physics and human behavior. According to Newtonian physics, vehicles with higher speeds have more kinetic energy to transfer to another person, vehicle, or object. This transfer of energy during a crash is the root cause of traffic injuries and fatalities, and this kinetic energy transfer can cause serious damage when a crash involves a bicyclist or a pedestrian. The higher the impact speed of a crash, the greater the risk of serious injury or death (Figure 1).

Figure 1. Risk of pedestrian death in relation to impact speed

Infographic detailing 5 scenarios where a car strikes a pedestrian at different speeds (20, 30, 40, 50 and 60 miles per hour) and the resulting fatality rates (8, 20, 46, 75 and 92 percent respectively).

Source: Data from Tefft 2013; Figure modified from:

In terms of human behavior, higher speeds increase the amount of information that drivers must process, leading to more driver stress and fatigue. Higher speeds also narrow drivers’ field of vision and require greater distances for drivers to react and come to a complete stop (Figures 2 and 3). Speed limit changes that slow down cars can reduce the reaction and braking distances needed to safely yield to pedestrians and bicyclists (NHTSA, 2015).

Figure 2. As speed increases, peripheral vision decreases

Figure 3: As speed increases, stopping distance increases

Below are summaries of evidence-based findings on the relationship between speed limits and safety outcomes, as well as areas where research is ongoing or unclear:

Main Takeaways

  1. A 5 mph reduction in speed limit is likely to decrease mean vehicle speed by 1-2 mph, or by 3 mph with stronger enforcement (Elvik et al. 2019). It may also reduce the speed of the fastest drivers to a much greater extent (Silvano & Bang 2016). The overall effects of a 5 mph speed limit reduction can lead to a 10-30% reduction in all fatalities and 2-15% reduction in serious bicyclist injuries (Elvik 2009; Helak et al. 2017; Zahabi et al. 2011).
  2. The risk of pedestrian fatality increases rapidly after 20 mph. At an impact speed of 30 mph, the risk of fatality is 20%, at 40 mph the risk of fatality is 46%, at 50 mph the risk of fatality is 75%, and at 60 mph the risk of fatality is almost certain, at 92% (Tefft 2013). Roads with a 5 mph lower speed are associated with 56-88% fewer serious pedestrian injuries and 80-96% fewer pedestrian fatalities (Hussain et al. 2019). It is important to clarify that this study examined impact speed rather than speed limits, though the two are related.
  3. Lower vehicle speeds may lead to improved safety perception with more people choosing to walk and bike, which in turn may lead to lower pedestrian and bicyclist crash rates. This phenomenon of “safety in numbers” for pedestrians and bicyclists is a continually evolving field of study (Elvik & Bjørnskau 2017). 

The following diagram shows how the posted speed limit can affect traffic characteristics, which in turn affect safety outcomes. Research outcomes on each of these links are summarized below.

Figure 4. Paths by which speed limits can affect safety outcomes

Flow chart detailing how speed limits affect traffic characteristics and safety outcomes. For more information, please see the following summary.

A (Speed Limit & Traffic Speed) : How does the speed limit affect traffic speed? 

Research has shown that reducing the speed limit by 5 mph typically results in a decrease in mean speed of 1-2 mph, or by 3 mph with stronger enforcement (equivalent to 60% of the speed limit change) (Grembek et al. 2020; Elvik 2009; Silvano & Bang 2016; Islam, El-Basyouny, Ibrahim 2014). This finding is consistent for limited access roadways as well as complete streets with multiple types of road users. Lowering speed limits can also reduce the speed of the fastest drivers (Silvano & Bang 2016; Grembek et al. 2020).

A (Speed Limit & Traffic Speed) + D (Traffic Speed & Likelihood of Crash): How does the speed limit affect the likelihood of a crash? 

Most studies suggest that higher speed limits are correlated with higher numbers of crashes. However, this positive relationship between speed limits and crash incidence is not consistent (Grembek et al. 2020, 20), and the variation across study findings makes it difficult to estimate the precise effect of traffic speed on crash frequency. For example, one study found that every 1 mph reduction in average speed can lead to a decrease of 2-7% in crash frequency (Taylor, Lynam, Baruya 2000). Another study found that a 1 mph reduction in speed at around 20 mph can lead to a decrease of around 12% in crash frequency (Elvik, Christensen, Amundsen 2004).

C (Impact Speed & Crash Severity): How does the impact speed affect crash severity? 

Research has consistently shown that reducing speeds on roadways is associated with fewer injuries, particularly fatal ones (Grembek et al. 2020, 8). Figure 5 depicts the results of a study on pedestrian risk using data from the National Highway Traffic Safety Administration (NHTSA). The study found that the risk of pedestrian fatality increases rapidly after 20 mph. At an impact speed of 30 mph, the risk of serious injury is almost 50% and the risk of fatality is more than 20%. Furthermore, the risks of serious injury or death are higher for older pedestrians and for those hit by light trucks as opposed to cars (Tefft 2013). 

Importantly, the data in this study that are shown in Figure 5 are based on crashes that occurred in the United States in the years 1994 – 1998, when average vehicle weights and shapes were substantially less dangerous than the present-day average vehicle. Because of the age of the data, the trends of increasing vehicle size and weight, and the relationship between heavier vehicles and vehicles with higher front ends and pedestrian fatality, these estimates are likely to be lower than what might be seen if this study were replicated today (Li 2012; Tyndall 2023).

Figure 5. Risk of serious injury and death in relation to impact speed

Graph detailing two charts that plot Impact Speed with Risk of Severe Injury and Risk of Death respectively.

Source: Tefft 2013

Figure 5 Caption: Risk of severe injury (left) and death (right) in relation to impact speed in a sample of 422 pedestrians aged 15+ struck by a single forward-moving car or light truck, United States, 1994-1998. Risks are adjusted for pedestrian age, height, weight, body mass index, and type of striking vehicle, and standardized to the distribution of pedestrian age and type of striking vehicle for pedestrians struck in the United States in years 2007-2009. Dotted lines pointwise 95% confidence intervals. Serious Injury is defined as AIS score of 4 or greater and includes death irrespective of AIS score. 

A (Speed Limit & Traffic Speed) + B (Traffic Speed & Impact of Speed) + C (Impact Speed & Crash Severity) and A (Speed Limit & Traffic Speed) + D (Traffic Speed  & Likelihood of Crash) : How does the speed limit affect the number of injuries and fatalities? 

The bulk of research indicates that reducing the posted speed limit by 5 mph, under certain circumstances, can lead to an 8-15% decrease in injuries and a 10-30% decrease in fatalities (Grembek et al. 2020; Elvik et al. 2019), with the strongest effect for middle-range speeds. Some outlier studies have reported reductions in injuries as high as 28% and 39% stemming from a 5 mph reduction in the posted speed limit (Elvik et al. 2019). Using past empirical evidence, Elvik (2019) concluded that injuries and fatalities relate to speeds exponentially to the fourth power (Elvik et al. 2019). While the exact result of speed limit reduction on injury and fatality frequency can vary with road type, the original posted speed limit, and urban context (Alnawmasi & Mannering 2022), there is consistent evidence that even modest decreases in average speeds on roadways could have a significant impact on the incidence of serious injury and fatality crashes. One strong example comes from a study of rural roads in Sweden, which found not only that reducing the posted speed limit led to fewer fatalities, but that increasing it led to more fatalities (Vadeby & Forsman 2018). Another study out of British Columbia, Canada found that increases in speed limits on rural highways lead to an increase in overall crashes and a large increase in the number of fatal crashes (Brubacher et al. 2018). 

While it is challenging to isolate the effects of speed limits on safety outcomes, especially for pedestrians and bicyclists, it is abundantly clear that roads with lower speeds pose a lower safety risk for pedestrians and bicyclists. One review of multiple studies found that a 5 mph decrease in vehicle speeds involved in a crash is associated with 56-88% fewer serious pedestrian injuries and 80-96% fewer pedestrian fatalities (Hussain et al. 2019).For bicyclists, research suggests that reducing the posted speed limit from 35 mph to 30 mph can reduce injuries to bicyclists by 17-32% and fatalities by 21-45% (Grembek et al. 2020).

4. Federal and State Guidelines on Safe Speeds for Local Governments

There is growing consensus among academic researchers and across various levels of government around using the Safe System and Safe Speeds approach to set speed limits to address traffic safety. Local jurisdictions in California have been constrained by state law and policy guidelines that require the use of the 85th percentile speed as the first step in setting speed limits in many instances. While the 85th percentile speed continues to be a starting point for speed limit setting in California, recent state legislation expands upon the existing methodology in speed limit setting by giving local jurisdictions more factors to consider in addition to the 85th percentile when setting speed limits.  

Local vs. State Control of Roadways

An important component of the speed limit setting process is determining which entity is responsible for the road under consideration. Roads are generally classified as locally-controlled roads or state-controlled roads. State-controlled roads, or roads on the “State Highway System” are generally highways and other arterials that have relatively higher speed limits and carry traffic between multiple municipalities. State-controlled roads are generally managed by the statewide department of transportation, which is Caltrans in California. In contrast, locally-controlled roads generally tend to carry relatively smaller volumes of traffic when compared to state highways and tend to begin and end within the boundaries of the governing jurisdiction. While state-controlled roads generally tend to carry larger volumes of traffic, there is a wide range of road types that qualify as “locally-controlled roads,” and these can range from larger arterials, to mid-sized collectors, to the smallest volume roads which are called “local roads.” Every California road designation can be viewed on the Caltrans California Road System Map

Responsibilities for maintenance and operations on certain segments of state-controlled roads may, in some cases, be relinquished to local jurisdictions. This is more commonly seen on state-controlled roads that cross through large cities such as Los Angeles or Sacramento. Occasionally, state-controlled roads may be referred to as “county roads” if a county is responsible for managing the road or a section of it. While certain segments of state-controlled roads may be colloquially referred to as “local roads” by residents if the state has relinquished responsibility over that section to a local jurisdiction, we will refer to these sections of state-owned roads under local governance as “locally-maintained state roads” rather than “local roads” to enhance clarity. 

Local Speed Limit Setting in California & Recent Legislation 

Local jurisdictions in California are generally constrained in their ability to set speed limits by state laws and regulations, including provisions in the California Vehicle Code, California Manual on Uniform Traffic Devices (California Department of Transportation 2023), and guidance from the California Manual for Setting Speed Limits (CMSSL 2020). It is important to note that, as of December 2023, the CMSSL has not yet been updated to incorporate the latest legislation pertaining to speed limit setting (such as AB 43 and AB 1938) but still provides important information about the speed limit setting process. 

The requirements and guidance across the CVC, CA MUTCD and CMSSL include context-specific prima facie speed limits, Engineering and Traffic Study (E&TS) procedures, allowable deviations from the 85th percentile speed, signage requirements, and radar enforceability. An important exception to these requirements relate to local roads (as defined under the California Road System maps), which do not require an E&TS to designate a speed limit and are assigned speed limits deemed appropriate by local transportation engineers.

In addition to local roads, which do not require an E&TS, some corridors qualify for prima facie speed limits, which are speed limits that are designated by law for certain types of roadways (e.g., alleys, school zones, senior zones). Many of these prima facie speed limits apply even in the absence of actual speed limit signage. Some roads have a prima facie speed limit of 25 mph (such as those in senior zones, school zones and business activity districts), and are therefore excluded from certain speed trap provisions (CVC § 40802). 

Outside of local roads and prima facie zones, opportunities to comprehensively lower speed limits without an 85th percentile speed-based E&TS are limited. Importantly, roads that carry higher volumes, such as arterial or collector roads, still do require speed limits based on the 85th percentile speed through an E&TS when no prima facie speed is designated. These high-volume roadways in urban settings are more likely to be on a jurisdiction’s High Injury Network due to higher serious injury and fatal crash incidence. Local jurisdictions may find safe speed limit setting and protecting vulnerable users especially challenging on these roadways when prevailing driver speeds are higher than is safe for all roadway users. 

Recent state legislation (AB 43, AB 1938 and AB 321) has given local jurisdictions more flexibility in setting speed limits on their locally-controlled roads in three major ways: 

  1. The creation of new prima facie zone legislation allowing lower speed limit designations in business activity districts and certain school zones.

  2. Granting local jurisdictions the ability to further lower speed limits in certain safety-related settings relative to 85th percentile speeds measured during the E&TS process. 

  3. Granting local jurisdictions the ability to retain current speed limits or restore the immediately prior speed limit on a corridor, especially if it was raised in the past 10 years, even if the current E&TS shows a higher 85th percentile speed. 

Federal Policy Direction 

In 2022, the US Department of Transportation’s National Roadway Safety Strategy (USDOT 2022) adopted the Safe System approach as a guiding paradigm to address traffic fatalities and serious injuries nationwide. One of the action areas led by the Federal Highway Administration (FHWA) is to “promote safer speeds for all users through context-appropriate speed limits, road designs, and other practices” (USDOT 2022).

Currently, the CA MUTCD recommends using “the 85th percentile rule” to set speed limits but is in the process of being updated to reflect changes allowed by recent state legislation that give local jurisdictions more flexibility to deviate from the 85th percentile speed. This is critical because the National Transportation Safety Board (NTSB) found that relying on the 85th percentile speed to change speed limits in high-speed zones results in “higher operating speeds and new, higher 85th percentiles in the speed zones, and an increase in operating speeds outside the speed zones” (Grembek et al. 2020, 54; California Department of Transportation 2023; NTSB 2017). The NTSB recommended incorporating the Safe System approach for urban roads to strengthen protection for vulnerable users (NTSB 2017).

A new 11th Edition of the Federal MUTCD was issued in December 2023. It expanded upon the approach for setting speed limits, emphasizing the need to consider the roadway context, especially in urban areas where the 85th percentile speed is higher than the desired speed limit due to the presence of vulnerable road users. This new edition downgraded many of the guidelines (“should” statements in the prior version) to “support” statements (i.e., information to consider). The Federal MUTCD does not have force or authority in California. The State is granted two years to adopt the Federal MUTCD or to amend the State MUTCD to be consistent with the Federal version. It is unknown how this process may unfold in California, but State law requirements for setting speed limits generally cannot be nullified by contrary provisions of the Federal MUTCD.

Caltrans Director’s Policy and the 2020-24 Strategic Highway Safety Plan 

In 2022, Caltrans issued Director’s Policy 36, which adopts the Safe System Approach as the framework for achieving the vision to eliminate fatalities and serious injuries on California’s roadways by 2050 (State of California 2022). This policy directs all Caltrans divisions to align their practices with and promote the implementation of the Safe System approach (State of California 2022). 

Caltrans’ 2020-24 Strategic Highway Safety Plan incorporates the Safe System approach as a guiding principle to addressing various High Priority Challenge Areas, including Speed Management/Aggressive Driving and Active Transportation Safety (California Department of Transportation 2023). Specifically, the Implementation Plan for Speed Management and Aggressive Driving called for implementing a new roadway-based, context-sensitive approach to establish speed limits in California that prioritizes the safety of all road users (California Department of Transportation 2023). 

Zero Traffic Fatalities Task Force 

In 2019, the California State Transportation Agency (CalSTA) established the Zero Traffic Fatalities Task Force (ZTFTF) under AB 2363 to develop policy recommendations for reducing statewide traffic fatalities to zero (California State Transportation Agency 2023). The task force assessed the current procedures for setting speed limits in California, which rely on “the 85th percentile rule,” and recommended greater flexibility for local jurisdictions to set speed limits. 

One of the ZTFTF’s long-term recommendations is to develop a new roadway-based context sensitive approach to establish speed limits that prioritizes the safety of all road users. In the meantime, the report also recommended various changes to speed limit setting, such as increasing the reduction allowance for posted speed limits to allow for greater deviations from the 85th percentile speed. Some of these recommendations have been incorporated into recent legislation that prioritizes multimodal safety and provides local jurisdictions with greater flexibility in setting speed limits on roads within their control and are currently in the process of being implemented into the upcoming newest edition of the CA MUTCD.