Request Free Consultation

An Analysis of Drug-Related Fatal Crashes in the United States

Request Free Consultation
September 13, 2021 Car Accidents

On June 4, 2019, a drugged driver with a suspended driver’s license was responsible for the death of one woman and injuring three others in Temecula, California. As he slammed into the back of a Nissan Altima at 100 mph, Javier Caldera started a six-car pileup. The intensity of the crash caused one vehicle to fly across the freeway and into a City Hall parking structure. Two years later, Caldera was deemed “an extreme danger to our community” and sentenced to over 23 years to life.

Even in this serious sentence, no disclosure was made about the type or amount of drug used by Caldera.

As recreational marijuana becomes legalized in more states, drugged driving has attracted more news coverage in recent years. However, data related to drugged driving is limited compared to drunk driving. Due to a lack of standards for drug testing and reporting throughout the United States, it is difficult to analyze the impact of drugged driving on fatal crashes.

Acknowledging these limitations, we worked with the data visualization agency, 1Point21 Interactive to analyze available drugged driving data over the past two decades in the United States. Our goal was to assess the impact of geographic location, social status, and changes in drug laws on fatal crashes.

[Interactive Map] Drug-Related Fatal Crashes from 2000 – 2019

For this study, we focused on drivers in fatal crashes who tested positive for drugs between 2000 and 2019. (Alcohol was not considered in this study). During this time frame, 13.5 percent – or 93,970 – of the total 691,605 fatal crashes involved one or more drivers who tested positive for at least one drug. Drivers tested positive for marijuana more frequently than any other drug, followed by amphetamines and opioids. Overall, the data shows that the number of drivers in fatal crashes who test positive for drugs is rising, though this doesn’t necessarily indicate that the drug caused the crash.

As mentioned earlier, drug testing and reporting across the U.S. isn’t standardized. Therefore, our data is incomplete and is impacted by the following factors:

  • How often states drug test drivers in fatal crashes
  • How accurate and complete the drug tests are. The more drugs someone is tested for, the more likely at least one result could be positive.

What Are the Most Common Drugs Involved in Fatal Crashes?

Marijuana was the most common drug involved in fatal crashes during this time period. Drivers in fatal crashes who tested positive for marijuana more than quadrupled from 2000 to 2019 (1,219 to 5,223).

Overall, twice as many drivers tested positive for marijuana as they did for other drugs, such as amphetamine, opioids, benzodiazepines, or cocaine. Drivers were the least likely to test positive for PCP/ Phencyclidine (angel dust), which is known for its mind-altering effects that can lead to hallucinations, as they were for “other” drugs, which include drugs such as inhalants, barbiturates, and various Rx pills.

Percent of Drivers Who Tested Positive for Indicated Drug from 2000-2019:

  • Marijuana: 6.8 percent
  • Amphetamines: 3.5 percent
  • Opioids: 3.5 percent
  • Benzodiazepines: 3.1 percent
  • Cocaine: 2.8 percent
  • Other depressants: 0.5 percent
  • Hallucinogens: 0.2 percent
  • Other stimulants: 0.2 percent
  • PCP: 0.1 percent
  • Other: 0.1 percent

Cocaine’s Decline

Interestingly, positive tests for cocaine in fatal crashes dipped from 2007 to 2010, along with the overall U.S. consumption of cocaine. According to a report produced for the White House Office of National Drug Control Policy, the amount of pure cocaine consumed in the United States declined by half between 2006 and 2010, from an estimate of 300 to 150 metric tons of pure cocaine. Though it isn’t clear what exactly caused this decline, other factors such as the production in Columbia, cocaine seizures, and the estimated amount of cocaine shipped to the U.S. also reflected similar decreases during the same time frame.

Opioid Crisis

The opioid crisis can also be seen in this data – specifically the second wave beginning in 2010. The opioid crisis was so significant that it impacted  U.S. life expectancy. Positive tests for opioids in fatal crashes peaked in 2015 – the same year life expectancy in the U.S. dropped for the first time since 1993.

Marijuana Drug Tests: Before and After Legalization

Have Positive Marijuana Drug Tests Increased in States that Legalized it?

Surprisingly, the number of positive marijuana tests in fatal crashes has not shown an increase in states that legalized marijuana. In fact, the number of positive marijuana tests is increasing in all states, regardless of whether or not marijuana has been legalized.

Colorado and Washington were the first states to legalize marijuana, followed by California. A study analyzed the changes in Colorado and Washington’s motor vehicle fatality crash rate three years after recreational marijuana was legalized and compared it to similar states that had not legalized marijuana. It concluded that drug-related fatal crashes did not increase in Colorado or Washington post-legalization, compared to the other states.

The legalization of marijuana in Colorado and Washington also shows minimal to no effect on traffic fatalities involving teen drivers. Some states are finding more impaired drivers, but only because they tend to monitor them more proactively. Colorado, for example, only began tracking levels of marijuana impairment in drivers in 2016.

How Do Drug Tests Work?

Administering Drug Tests

There isn’t a standard protocol around drug testing. A police officer will only test for drugs in a fatal crash if there is probable cause – an indication that impaired driving played some role in the accident.

Therefore, the number of positive drug tests varies based on the frequency of drug tests given in each state. Some states – like North Carolina – simply do not test their drivers very often, which is reflected in our data. In all the heatmaps, North Carolina shows up as white. This does not indicate that people in North Carolina do not do drugs and drive – just that the state doesn’t test for them.

Detecting Drugs

Due to differences in procedure, it also isn’t clear if the driver was at fault or the one to die in the crash. The drug test doesn’t indicate the time when the drug was taken, so the presence of a drug cannot be automatically correlated to impairment. For example, urine tests can detect marijuana in the urine for an estimate of 3 to 30 days after use. Saliva tests, however, can detect marijuana for about 24 to 72 hours after use. Hair tests can detect THC for up to 90 days after use.

Where Are the Most Common Positive Drug Tests by County?

The map shows that positive drug tests in fatal crashes – for any drug – appear to be the most common in counties in the following states: Washington, Montana, California, West Virginia, Tennessee, Ohio, New Hampshire, Vermont, and Kentucky. But it is also probable that crash-related drug testing in those counties is much more frequent.

Hotspots in the Country for Specific Drugs


From the map, we can see that positive tests for marijuana in fatal crashes might be more common in tourism spots – such as Hawaii, Alaska, Washington, and California. Alaska’s capital, Juneau, has the highest percentage of drivers in fatal crashes who tested positive for marijuana, and three counties in Hawaii are hotspots. Boulder, Colorado is also in the top ten hotspots.

Top 25 Counties With Highest Percent of Drivers Who Tested Positive for Marijuana (Population > 25k):

  1. Juneau, Alaska – 35 percent
  2. Stevens, Washington – 35 percent
  3. Mendocino, California – 33 percent
  4. Logan, Illinois – 31 percent
  5. Sullivan, New York – 31 percent
  6. Hawaii, Hawaii – 30 percent
  7. St. Louis, Missouri – 30 percent
  8. Boulder, Colorado – 29 percent
  9. Spokane, Washington – 28 percent
  10. Clallam, Washington – 27 percent
  11. Snohomish, Washington – 27 percent
  12. Kitsap, Washington – 26 percent
  13. Island, Washington – 26 percent
  14. Skagit, Washington – 26 percent
  15. Clermont, Ohio – 26 percent
  16. Pierce, Washington – 26 percent
  17. Matanuska-Susitna, Alaska – 25 percent
  18. Macoupin, Illinois – 25 percent
  19. Maui, Hawaii – 25 percent
  20. Thurston, Washington – 25 percent
  21. Santa Cruz, California – 24 percent
  22. Del Norte, California – 24 percent
  23. Grafton, New Hampshire – 24 percent
  24. Franklin, Illinois – 25 percent
  25. Orange, New York – 25 percent


The impact of the opioid crisis is seen in numerous counties in Kentucky, with the highest number of drivers in fatal crashes testing positive for opioids in Floyd, Kentucky. Interestingly, 21 out of the 25 hotspots are located in Kentucky, with the remaining in West Virginia or Vermont. In 2019, the University of Kentucky was awarded an $87 million grant to reduce opioid overdose deaths by 40 percent in the 16 counties that make up a third of Kentucky’s population.

Top 25 Counties With Highest Percent of Drivers Who Tested Positive for Opioids (Population > 25k): 

  1. Floyd, Kentucky – 46 percent
  2. Bell, Kentucky – 39 percent
  3. Pike, Kentucky – 38 percent
  4. Harlan, Kentucky – 36 percent
  5. Perry, Kentucky – 27 percent
  6. Knox, Kentucky – 25 percent
  7. Montgomery, Kentucky – 24 percent
  8. Boyd, Kentucky – 24 percent
  9. Boone, Kentucky – 23 percent
  10. Graves, Kentucky – 21 percent
  11. Pulaski, Kentucky – 21 percent
  12. Jessamine, Kentucky – 20 percent
  13. Logan, West Virginia – 19 percent
  14. Orleans, Vermont – 19 percent
  15. Mingo, West Virginia – 19 percent
  16. Madison, Kentucky – 19 percent
  17. Laurel, Kentucky – 19 percent
  18. Mercer, West Virginia – 18 percent
  19. Hopkins, Kentucky – 18 percent
  20. Marshall, Kentucky – 18 percent
  21. Greenup, Kentucky – 18 percent
  22. Grayson, Kentucky – 18 percent
  23. Roane, Kentucky – 17 percent
  24. Whitley, Kentucky – 17 percent
  25. Kenton, Kentucky – 17 percent


Multiple counties within West Virginia, Illinois, and Ohio are hotspots. Maryland’s St. Mary’s had the highest percentage of drivers who tested positive for cocaine in fatal crashes. St. Mary’s has a known history of cocaine distribution and use, and in 2018, its cocaine ring leaders were sentenced to over 14 years in prison.

Top 25 Counties With Highest Percent of Drivers Who Tested Positive for Cocaine (Population > 25k): 

  1. St. Mary’s, Maryland – 13 percent
  2. Vermilion, Illinois – 11 percent
  3. Macon, Illinois –11 percent
  4. Sullivan, New York – 10 percent
  5. Rio Arriba, New Mexico – 10 percent
  6. Charlottesville, Virginia – 10 percent
  7. Berkeley, West Virginia – 10 percent
  8. Santa Cruz, Arizona – 10 percent
  9. St. Louis, Missouri – 10 percent
  10. Wakulla, Florida – 10 percent
  11. Franklin, Ohio – 9 percent
  12. Jones, Georgia – 9 percent
  13. St. Charles, Louisiana – 9 percent
  14. Coffee, Georgia – 9 percent
  15. Clarke, Georgia – 9 percent
  16. Skagit, Washington – 9 percent
  17. Bristol, Rhode Island – 9 percent
  18. Montgomery, Ohio – 9 percent
  19. Champaign, Illinois – 9 percent
  20. Peoria, Illinois – 9 percent
  21. Cuyahoga, Ohio – 9 percent
  22. Madison, Ohio – 9 percent
  23. Winnebago, Illinois – 9 percent
  24. LaSalle, Illinois – 8 percent
  25. Milwaukee, Wisconsin – 8 percent


For the most part, positive tests for benzodiazepines in fatal crashes seem to be concentrated in Kentucky, Tennessee, Ohio, and West Virginia. Nine counties in Tennessee are hotspots, including Lawrence, Tennessee – the county with the highest percentage of drivers who tested positive for benzodiazepines in fatal crashes during our study period. In response to the misuse of drugs, in 2013, Tennessee passed a law restricting the quantities of prescription drugs for opioid pain medicines and benzodiazepine medicines that can be dispensed. Months prior to this, medical professionals in Tennessee were instructed to check the Controlled Substance Monitoring Database before prescribing, to identify a patient’s drug dependence.

Top 25 Counties With Highest Percent of Drivers Who Tested Positive for Benzodiazepines (Population > 25k): 

  1. Lawrence, Tennessee – 27 percent
  2. Pike, Kentucky – 26 percent
  3. Floyd, Kentucky – 25 percent
  4. Roane, Tennessee – 25 percent
  5. Bell, Kentucky – 22 percent
  6. Perry, Kentucky – 21 percent
  7. Jefferson, Tennessee – 21 percent
  8. Whitley, Kentucky – 20 percent
  9. Fayette, West Virginia – 20 percent
  10. Hamblen, Tennessee – 20 percent
  11. Obion, Tennessee – 20 percent
  12. Boyd, Kentucky – 20 percent
  13. Hopkins, Kentucky – 19 percent
  14. Greene, Tennessee – 19 percent
  15. Hart, Georgia – 19 percent
  16. Lincoln, Tennessee – 18 percent
  17. Logan, West Virginia – 18 percent
  18. Marshall, Kentucky – 18 percent
  19. Montgomery, Kentucky – 17 percent
  20. Graves, Kentucky – 17 percent
  21. Monroe, Tennessee – 17 percent
  22. Brown, Ohio – 17 percent
  23. Mingo, West Virginia – 17 percent
  24. Franklin, Missouri – 17 percent
  25. Cocke, Tennessee – 16 percent


Positive tests for amphetamines in fatal crashes are concentrated in Hawaii and in the northwestern states – including California, Oregon, and Washington. Three counties in Hawaii are hotspots – just as many as for marijuana. It seems that Hawaii’s “ice” (crystal meth) epidemic from the 1980s never really went away but has left a lasting impression on the community.

Top 25 Counties With Highest Percent of Drivers Who Tested Positive for Amphetamines (Population > 25k): 

  1. Glenn, California – 28 percent
  2. Island, Washington – 21 percent
  3. Tuolumne, California – 20 percent
  4. Douglas, Washington – 20 percent
  5. Hawaii, Hawaii – 19 percent
  6. Amador, California – 17 percent
  7. Lyon, Nevada – 17 percent
  8. Mendocino, California – 17 percent
  9. Grays Harbor, Washington – 16 percent
  10. Butte, California – 16 percent
  11. Clark, Washington – 16 percent
  12. Calaveras, California – 15 percent
  13. Kauai, Hawaii – 15 percent
  14. Maui, Hawaii – 15 percent
  15. El Dorado, California – 14 percent
  16. Tulare, California – 14 percent
  17. Crawford, Arkansas – 14 percent
  18. White, Arkansas – 14 percent
  19. Lake, California – 14 percent
  20. Kings, California – 14 percent
  21. Cleburne, Arkansas – 14 percent
  22. Pontotoc, Oklahoma – 14 percent
  23. Tehama, California – 14 percent
  24. Columbia, Oregon – 14 percent
  25. Lassen, California – 14 percent

Preventing Drugged Driving

Like drunk driving, drugged driving – in any amount – impairs judgment, coordination, and reaction time. Different drugs can affect people differently, and it is difficult to predict just how much you will be impacted. Recognize that even though recreational marijuana is legal in certain states, it can still be a threat to you and others on the road. Studies show that marijuana can negatively impact drivers – causing more lane weaving, lowering reaction time, and altering attention on the road. Cocaine or methamphetamine can result in aggressive and reckless driving, and prescription drugs can make you feel drowsy, dizzy or lacking in judgment. This is why the best way to protect yourself and others is to avoid driving under the influence of drugs in any circumstance. Stay safe by practicing the following tips:

  • Don’t risk it; always plan for alternative transportation, whether that is calling a friend, family member, or rideshare ahead of time.
  • Don’t let others risk it; stop your friend or loved one from driving under the influence of drugs.
    • Be mindful of the level of impairment of the other person, and speak slowly.
    • Make it clear that you are suggesting alternative transportation because you care and want to keep your loved one safe.

Methodology and Fair Use

For this study, we analyzed 2000-2019 fatal crash data from the National Highway Traffic Safety Administration (NHTSA). This data only includes drivers who tested positive for drug tests. This does not include drivers who tested positive for alcohol or who were simply not given drug tests at all following the fatal crash.

If you would like to report or republish our findings, please link to this page to provide a citation for our work.