Skip to main content
You can also sort pages by filters.
Table of Contents
Download the Full Book

Compton et al. (2009) describe four basic issues that must be addressed to better understand the extent of the problem of drug-impaired driving:

  • What drugs impair driving ability?
  • What drug dose levels are associated with impaired driving?
  • How frequently are impairing drugs being used by drivers?
  • What drugs are associated with higher crash rates?

There are still sizeable gaps in our understanding of the effects of drugs on driving. In their review of drug-impaired driving, Jones et al. (2003) concluded: “The role of drugs as a causal factor in traffic crashes involving drug-positive drivers is still not understood…. Current research does not enable one to predict with confidence whether a driver testing positive for a drug, even at some measured level of concentration, was actually impaired by that drug at the time of crash.” (p. 96).

Despite these challenges, a growing body of research suggests that some over-the-counter prescription medications, and illegal drugs may impair a driver’s ability to operate a vehicle (for reviews, see Couper & Logan, 2004; Gjerde et al, 2019; Jones et al., 2003; Kelly et al., 2004; and Strand et al., 2016). Much of this research has involved laboratory or experimental studies using driving simulators. Some epidemiological studies have examined the effect of drugs on crash incidence and crash risk. See Compton et al. (2009) and Compton and Berning (2015) for a discussion of this research.

In most cases, the research investigating the effect of drugs on driving has had variable results, depending on the drug examined and the methodology employed. In addition, for many drugs, there is a lack of high-quality, published research. A summary of suspected crash risks associated with specific classes of drugs are provided below, with research gaps acknowledged.

Cannabis (often referred to as marijuana)

Cannabis is a term for all products derived from the plant Cannabis sativa or Cannabis indica. While cannabis contains many unique chemical compounds (i.e., cannabinoids), the substance primarily responsible for the psychoactive effects of the drug is tetrahydrocannabinol (THC)[1] (NIDA, 2019). According to the National Institute on Drug Abuse (2019), the acute effects of smoking or consuming THC include: an altered sense of time, distorted perception, impaired coordination, memory loss, and difficulty problem-solving. Regarding the impacts of cannabis on driving, some simulator and on-road, experimental studies have found increased car-following distance, increased lane position variability (i.e., weaving), increased driver reaction time, and decreased performance on divided attention tasks (e.g., performing two or more subtasks simultaneously) (Hartman & Huestis, 2013; Pearlson et al., 2021; Sewell et al., 2009); however, for these and other performance measures, some studies have found an association, while others have found no effects or inconclusive results (Pearlson et al., 2021). Caution should be used when interpreting the findings of simulator and experimental studies, because both types of studies may be susceptible to limitations, including an unrealistic driving environment, tightly controlled conditions, and a low number of study participants, among other factors. In addition, since these studies often examine many performance measures, significant results may be highlighted over nonsignificant results (Smiley, 1999). Also, unlike comparable studies involving alcohol, several studies examining the relationship between cannabis and driver impairment noted that study participants reported being aware of potential driving deficits related to their cannabis consumption, with some participants responding to these real or perceived deficits through compensatory behaviors, such as reducing speed (Pearlson et al., 2021). One finding that has been consistent across most simulator and on-road, experimental studies is that, unlike alcohol, there is little evidence to support a direct dose-response relationship between cannabis consumption, blood THC concentration, and driver impairment (Peterman, 2019).

The epidemiologic research examining the association between cannabis consumption and elevated real-world crash risk is also inconclusive. A large-scale study in Virginia Beach, Virginia, found no elevated crash risk for THC-positive drivers after adjusting for demographic variables and alcohol use (Compton & Berning, 2015). In addition, the Virginia Beach case-control study did not find evidence of a synergistic effect on crash risk among drivers testing positive for both alcohol and THC (Lacey et al., 2016). Conversely, a 2021 review of recently published meta-analyses and culpability studies found a slight, but significant, elevated crash risk after recent cannabis usage, especially among drivers with high blood THC concentrations. However, there was considerable heterogeneity across the reviewed studies in terms of rigor, design, and measurement of exposure and outcomes (Pruess et al., 2021). Nevertheless, with more than half of all States permitting medical or recreational cannabis usage, SHSOs are concerned about the potential traffic safety impacts of legalization (NCSL, 2022). While the prevalence of cannabis usage tends to increase in States with legalized cannabis, legalization’s impact on traffic safety outcomes is inconclusive (Zvonarev et al., 2019). A study examining the relationship between cannabis legalization and traffic fatalities in Colorado and Washington did not find an association (Hansen et al., 2018). Lane and Hall (2019) observed an increase in traffic fatalities in 2 out of 3 States after cannabis legalization (Colorado and Washington, with Oregon being the exception). However, the effect was modest (one additional fatality per million residents) and was followed by a trend reduction in traffic fatalities. The study of the relationship between cannabis legalization and fatal traffic crash risk is further complicated by challenges related to data quality and completeness at the national level (see Drug-Impaired Driving – Data/Surveillance).


There have been fewer studies examining the risks of stimulants (e.g., amphetamines) on driving. A meta-analysis performed by Elvik (2013) found elevated odds of fatal crash involvement among drivers under the influence of amphetamines as well as cocaine. However, causal relationships could not be established due to the paucity of high-quality studies available, with many of the studies inadequately controlling for potential confounders, among other limitations (Elvik, 2013). Therefore, more research is needed to better characterize the direction and magnitude of relationships between specific categories of stimulants and motor vehicle crash risks.


The relationship between antihistamines and motor vehicle crashes is ambiguous (Moskowitz & Wilkinson, 2004). A small connection has been found between first-generation antihistamines (e.g., diphenhydramine) and crashes, but second-generation (non-sedating) antihistamines (e.g., cetirizine) do not appear to be associated with an elevated crash risk (Perttula et al., 2014). Extended use of antihistamines may also be associated with crash risk, but this is also an area that needs additional study (Gibson et al., 2009).


Like many of the other classes of drugs under discussion, results for antidepressants are mixed. The use of second-generation antidepressant medications, such as selective serotonin reuptake inhibitors, appear to have minimal or no relationship with crash risk, especially when a person is acclimated to the drug, but this is not necessarily the case with older types of antidepressants and antidepressants with sedative qualities, such as tricyclic antidepressants and trazodone (Brunnauer & Laux, 2017; Hansen et al., 2015; Myers, 2021). However, it should be noted, that many studies examining the hazard posed by antidepressants are complicated by the fact that depression, along with other mental health conditions, are independent risk factors for traffic crashes (Hill et al., 2017). In these cases, antidepressants may decrease crash risk by treating an underlying potentially impairing medical condition. For example, a New Zealand study found that drivers with a history of suicidal ideation had an elevated risk of crash injury; however, this relationship was not significant among drivers reporting current antidepressant medication (Lam et al., 2005). More research is needed to better understand the relationship between depression (and other medical conditions) and crash risk, as well as the potential mitigating effects of antidepressants and other medications (Hill et al., 2017; Rapoport et al., 2022).

Narcotic Analgesics/Opioids

Opioids are a class of drug that includes both legal and illegal drugs, such as heroin, fentanyl, and prescription medications used for the treatment of pain, such as oxycodone (OxyContin), oxymorphone (Opana), and hydrocodone (Vicodin). Opioids bind to and activate opioid receptors in the human body, blocking pain and releasing dopamine, which can induce feelings of contentedness and relaxation. For this reason, prescription opioid analgesics have been used for decades for the treatment of acute and chronic pain, among other medical reasons. However, both licit and illicit opioids can cause sedation, confusion, slowed breathing, unconsciousness, and death (NIDA, 2021). Due to their sedative and other psychoactive effects, both licit and illicit opioids have been theorized to impair driving ability and result in negative traffic safety outcomes (Beaulieu et al., 2022). However, few experimental or epidemiological studies have adequately characterized the transportation safety risks associated with opioids (Beaulieu et al., 2022; Cameron-Burr et al., 2021; Leung, 2011).

In a recent clinical review, Beaulieu et al. (2022) found that most experimental studies indicated some level of impairment on tests of psychomotor function, including tests assessing driving performance; however, many of these studies suffered from the limitations common to experimental studies, as discussed previously. In addition, most study populations consisted of healthy people with no history of opioid use, a population that may not be representative of people who use opioids on a routine basis, such as people prescribed opioid analgesics for chronic pain management. In addition, a substantial number of studies found no evidence or inconclusive evidence of impairment. A 2011 study (Leung) found even more limited evidence supporting a relationship between opioid analgesics and impairment, especially at therapeutic doses. However, the authors noted that more research is needed, especially regarding the combination of opioid analgesics with other CNS depressants, such as alcohol.

To date, most of the epidemiologic literature is descriptive in nature, documenting the prevalence of opioid positivity in specific populations (e.g., people with serious or fatal motor vehicle crash injuries) (Beaulieu et al., 2022; Cameron-Burr et al., 2021; Leung, 2011; Thomas et al., 2022b).While several studies have noted an increase in the prevalence of drivers testing positive for various opioids, it is unclear how changes in testing procedures may be affecting trends (Cameron-Burr et al., 2021). Among studies examining the association between opioid use and motor vehicle crash risk, the evidence is mixed. While several recent epidemiologic studies have found an association between the use of opioids and motor vehicle crashes, other studies have failed to find an association (Gjerde et al., 2015). NHTSA’s Virginia Beach case-control study examined the crash risk of opioid analgesics, along with other prescription and illicit drugs. Narcotic analgesics were not found to be significantly associated with crash risk alone or in combination with alcohol; however, the number of participants testing positive for narcotic analgesics was not sufficient for stratifying analyses by concentration levels (Lacey et al., 2016).

Central Nervous System (CNS) Depressants

CNS depressants are prescription medications that include sedatives, tranquilizers, and hypnotics. These drugs tend to slow brain activity and produce quietening effects and are therefore used for treating sleep and anxiety disorders, among other conditions. CNS depressants include benzodiazepines (e.g., diazepam [Valium], alprazolam [Xanax], clonazepam [Klonopin]), non-benzodiazepine sedative hypnotics (e.g., zolpidem [Ambien], carisoprodol [Soma], eszopiclone [Lunesta]), and barbiturates. CNS depressants produce numerous effects including drowsiness, lack of coordination, difficulty concentrating, and confusion (NIDA, 2018). Barbiturates are now uncommon. However, benzodiazepines and non-benzodiazepine sedatives are a concern for driving because of their potentially impairing effects (Chong et al., 2013; Maust et al., 2019). Two literature reviews found that under experimental conditions, benzodiazepines adversely affect driving ability, particularly maintaining lateral position (Dassanayake et al., 2011; Verster & Roth, 2013). Regarding the epidemiologic evidence regarding CNS depressant use and crash risk, two meta-analyses found an association between benzodiazepine use and increased crash risk; however, neither study formally assessed study quality or publication bias (Dassanayake et al., 2011; Elvik, 2013; Rapoport et al., 2009). A more rigorous meta-analysis performed by Elvik (2013) found an elevated crash risk for benzodiazepines for property damage only, nonfatal injury, and fatal traffic crashes and an elevated property damage only (PDO) crash risk associated with the use of non-benzodiazepine sedative hypnotics. However, Elvik (2013) noted that due to heterogeneity in study design and the failure of many studies to adequately control for confounding factors, a causal relationship could not be established conclusively. Also, NHTSA’s Virginia Beach case-control study did not find a relationship between CNS depressants and crash risk with or without combination with alcohol (Lacey et al., 2016). Among studies finding an association between use of CNS depressants and crashes, the risk was modulated by the type of benzodiazepine used, the dose, the time elapsed since use, and whether the drug was combined with alcohol or other drugs, such as opioids (Dassanayake et al., 2011; Leung, 2011; Scherer et al., 2018).

Other Licit and Illicit Drugs

The preceding is not an exhaustive list of substances that may impair driving ability. Law enforcement officers commonly find drug impaired drivers on other substances such as ketamine, MDMA, and inhalants. Studies have suggested that antiemetic agents, antiepileptic agents, antiparkinsonian agents, antipsychotics, hypoglycemic agents, among other medications and drugs may negatively affect driving ability; however, for many of these agents, the evidence is insufficient (U.S. FDA, 2021; Hetland & Carr, 2014; Myers, 2021).

Like alcohol-impaired driving, drug-impaired driving is primarily addressed through a combination of laws, enforcement, and education (AAAFTS, 2018a; AAAFTS, 2018b). Relatively few countermeasures have been developed to specifically address drug-impaired, separate from alcohol-impaired driving, and there has been little evaluation of drug-impaired-driving countermeasures. AAAFTS investigated the potential for alcohol-impaired-driving countermeasures to be applied to drug-impaired driving. The conclusions point to the need for more research to better understand the nature and degree of traffic safety risk posed by drugs, as well as the effectiveness of potential countermeasures to address this issue. See the guide on drug-impaired driving produced by the Center for Problem-Oriented Policing for more information about drug-impaired-driving countermeasures (Kuhns, 2012). Cannabis-specific summaries can be found in NHTSA’s Marijuana-Impaired Driving: A Report To Congress (Compton, 2017) and the AAA Foundation for Traffic Safety’s report (Logan et al., 2016). Smith et al. (2018) reviewed the state of knowledge on countermeasures against impaired driving due to prescription and over-the-counter drugs.

[1] More specifically, the term THC usually refers to the delta-9-THC isomer (Δ9-tetrahydrocannabinol) (Felder & Glass, 1998).