IAFTC Newsletter. Volume 1. Issue 1. November 18, 2025.

Joshua Ott¹

¹Caselock, Inc., P.O. Box 285, Lebanon, GA 30146

This is an open-access article under the CC BY-NC-ND license.

Download PDF.

Abstract

The Drug Recognition Expert (DRE) program was developed to equip law enforcement officers with the ability to identify drivers impaired by drugs other than, or in addition to, alcohol. Although implemented across the United States, the program’s scientific basis remains contested. This paper critically examines the studies that validate the DRE’s 12-step Drug Influence Evaluation (DIE). A review of key studies—including the Johns Hopkins (1984), Los Angeles Police Department (1986), Arizona (1994), and Heishman et al. (1996, 1998) studies—reveals that DRE evaluations are not validated for actual impairment, but rather for the presence of drugs as confirmed by toxicology. False positive rates ranging from 5% to over 60% raise serious concerns about the accuracy of DRE assessments, particularly given their role in criminal prosecutions. The findings indicate that without independent, performance-based validation, the DRE program lacks the foundation necessary to serve as a scientifically valid measure of impairment.

Introduction

Drug-impaired driving presents a significant challenge for law enforcement, the legal community, and public safety. In response to limitations of toxicological testing, which cannot determine behavioral impairment, the Drug Recognition Expert (DRE) program was developed to train officers to recognize impairment in drivers by utilizing the Drug Influence Evaluation (DIE). The DIE is a standardized 12-step protocol intended to determine whether the suspect is impaired, if so, whether the cause is medical or drug-related, and if drug-related, which category or categories of drugs are the likely cause of the impairment. However, the program’s validation studies have not established that the DIE can accurately determine impairment as defined by being “less safe” to operate a vehicle. This paper critically examines the DRE program’s foundation, methods, and validation studies, arguing that the DIE lacks scientific validation for its stated purpose and carries an unacceptable rate of false positives.

Overview of the DRE Program

A Drug Recognition Expert (DRE) is an officer who has been trained to identify people who are impaired by drugs other than or in addition to alcohol. They are trained to use a standardized, systematic 12-step Drug Influence Evaluation (DIE) to determine three things.

1. Is the subject impaired?

2. Is the impairment due to medical issues or drugs?

3. If it is drugs, what category/categories of drugs is/are most likely causing the impairment? (1)

The 12-step DIE contains the following: breath test, interview of the arresting officer, preliminary examination, eye examinations, divided attention tests, vital signs (pulse is checked three separate times), pupil size and reaction to light, muscle tone, check for injection sites, subject’s statements, opinion of the evaluator, and toxicological analysis. 

To become certified, a DRE student must have a 75% toxicological corroboration rate. This means that if a DRE opines that one category of drugs is causing the impairment, that category must be confirmed by toxicology. If the DRE opines that two categories are causing the impairment, at least one must be confirmed by toxicology. If they opine that three categories are causing the impairment, at least two must be confirmed by toxicology. In the (Shinar & Schechtman, 1998) Study(2), it stated that DREs should be encouraged to always list two drug categories. Listing two drug categories increases the likelihood of the DRE identifying the correct drug category with no downside. 

Primary Validation Issue

The main issue in the DRE program lies in its reliance on toxicological confirmation as a measure of validity. Toxicology can confirm the presence of a drug or its metabolites, but it does not indicate whether the individual was behaviorally impaired. The use of toxicology in validation studies to determine the accuracy of DRE’s fails to measure the DIE’s intended purpose: identifying drivers who are less safe to operate a vehicle. This circular validation approach, where toxicology results are used to support DRE opinions, and DRE opinions are then used to claim that positive toxicology findings indicate psychoactive drug effects, fails to provide independent validation of the DIE’s ability to identify drug impairment. 

The DRE Manual states that one of the reasons for the DIE being needed is: “chemical tests usually disclose only that the subject has used a particular drug recently. The chemical test usually does not indicate whether the drug is psychoactive at the present time. Thus, the DRE procedures are needed to establish that the subject not only has used the drug, but also that he or she is under the influence.” (2025 DRE 7-Day Participant Manual, Session 3, Page 5)

Review of Key Studies

The DRE training course teaches officers about three studies, but the primary focus is on the Johns Hopkins Laboratory Study (1984)(3) and the Los Angeles Field Validation Study (1985)(4). The Arizona Study (1994)(5) is only briefly mentioned in the manual.

Johns Hopkins Laboratory Study (1984)

The Johns Hopkins study, sponsored by the National Highway Traffic Safety Administration (NHTSA), involved 80 male volunteers aged 18–35. Only four DRE raters were involved. The false positive rate was 5%. This false positive rate is potentially misleading due to the volunteers being trained on the psychomotor tasks and subjective effect questionnaires that were used in the study ahead of time. If they did not show adequate performance, they were not accepted for participation in the study. This removed potential false positives ahead of time. Additionally, the DREs were free to inquire how the subjects felt, had they ever felt that way before, or had they ever used drugs that made them feel that way, etc. These questions possibly unblinded the DREs and may have influenced their opinions.

Los Angeles Police Department Field Study (1986)

The LAPD study involved a total of 219 suspects. Twenty-eight suspects did not provide a blood sample and were not included in the final data. There were 18 suspects who were determined not to be under the influence of drugs by the DREs during the Preliminary Exam (Step 3 of the DIE) and were released from custody. This left a total of 173 suspects. 

All 173 were believed to be under the influence of drugs by the DREs. One suspect had no drugs or alcohol detected, and 10 suspects had only alcohol detected. This means that after Step 3 of the DIE, the DRE’s accuracy increased 0%. 

Adding the 18 suspects who were determined not to be under the influence of drugs during the Preliminary Exam (Step 3) to the 11 suspects who tested negative for drugs is 29. The DREs incorrectly determined that 11 out of those 29 were drug-impaired. That is a false positive rate of 37.9%.

Even the study’s authors acknowledged that drug presence does not equate to impairment, noting, 

There is no way to determine objectively whether the suspects were actually too “impaired” to drive safely. The fact that drugs were found in a suspect's blood does not necessarily mean the suspect was too impaired to drive safely. Contrary to the case with alcohol, we do not know what quantity of a drug in blood implies impairment. Thus, this study can only determine whether a drug was present or absent from a suspect's blood when the DRE said the suspect was impaired by that drug.(p. 15). 

This study failed to validate the DIE for identifying drug impairment.

Arizona Study (1994)

The Arizona study analyzed 500 records from an established DRE program. The false positive rate was 61.7%, with 42 of 68 individuals with no drugs detected being incorrectly ruled impaired. 

Heishman et al. Laboratory Studies (1996, 1998)

These were laboratory studies that were financially supported by NHTSA. Both studies used certified DREs to conduct evaluations on volunteers. The volunteers had a history of drug use and were dosed with a drug from the category that they had a history of using. In the 1996 study, DREs incorrectly identified impairment in 40.7% of placebo-dosed cases and were only 50.6% accurate in predicting the correct drug category. 

If the cases in which the DRE concluded that ethanol was causing the impairment were excluded, the DREs were only 44.4% correct in predicting the drug category. The issue with this study is that the volunteers did not reside at the location between study sessions. This allowed them to use drugs on their own and created the possibility that they were still impaired by drugs when they showed up to the next session and received a placebo dose.

The 1998 study addressed this issue by having the volunteers reside in a closed unit of the Addiction Research Center. The DREs incorrectly identified 28.1% of placebo-dosed participants as impaired and were only 32.1% accurate in identifying the correct drug category. These findings demonstrate that DREs frequently misidentify sober individuals as being impaired and lack consistent accuracy in identifying drug categories.

The (1996) study also noted, 

Until a broad range of drugs and drug doses are tested on the DEC evaluation and independent performance tests under laboratory conditions, it is difficult to assess the validity of the DEC evaluation with respect to impairment criteria. Such validation is critically needed, however, because the current means of confirming a DRE’s prediction of impairment is the presence of parent drug or metabolite in blood or urine, which, with the exception of ethanol, provides little, if any, information concerning behavioral impairment.

In the almost 30 years since this paper was published, there have been no known attempts to validate the DIE with respect to impairment criteria.

Study	Year	False Positive Rate (%)
Johns Hopkins	1984	5
LAPD	1986	37.9
Arizona	1994	61.7
Heishman et al.	1996	40.7
Heishman et al.	1998	28.1
Shinar & Schechtman	1998	56.9

Discussion

Through decades of research, no study has validated the DIE for determining actual impairment. Instead, the DIE’s accuracy is judged using toxicological confirmation, a method that is incapable of measuring drug impairment. False positive rates ranging from 5% to over 60% demonstrate an alarming lack of accuracy in correctly identifying drug-free subjects as non-impaired. 

This is particularly alarming in a world in which multiple states have legalized recreational marijuana and many people are prescribed medications that fall into the CNS Depressant, CNS Stimulant, and Narcotic Analgesic categories of the DRE program. It is very concerning in a context where DRE opinions can significantly influence legal outcomes. The omission of critical research, such as the Heishman et al. studies, from current DRE training materials raises many concerns. Without testing the accuracy of the DIE with independent performance tests as cited in Heishman et al. (1996), the DIE cannot be considered a scientifically validated method of detecting drug impairment.

Conclusion and Recommendations

The DRE program’s reliance on toxicological corroboration instead of behavioral impairment undermines its scientific legitimacy. Given the serious legal and social implications of incorrectly identifying a sober person as impaired, the DRE program’s continued use without independent validation is not acceptable. Future research must address two foundational questions: 

1. Can DREs accurately identify drivers who are less safe to operate a vehicle due to drug impairment? 

2. Can DREs accurately discriminate sober individuals from impaired ones? 

Until such evidence exists, DRE evaluations should be handled with significant caution in the legal community.

AI Use Disclosure

The author acknowledges the use of ChatGPT to refine sentence structure, enhance clarity, and correct grammatical errors. All substantive content and conclusions are the author’s own.

Conflict of Interest

The author is a consultant and expert witness for DUI cases.

References

1. 2025 Drug Recognition Expert Manuals (Pre-School and 7-Day School Instructor and Participant Manuals)

2. Modeling the DRE Evaluation of Signs and Symptoms to Improve the Validity of Drug Impairment Diagnosis, David Shinar and Edna Schechtman (1998)

3. Identifying Types of Drug Intoxication: Laboratory Evaluation of a Subject-Examination Procedure, Bigelow, Bickel, Roache, Liebson, Nowowieski (Johns Hopkins -1985)

4. Field Evaluation of the Los Angeles Police Department Drug Detection Procedure, Compton, (LAPD -1986)

5. Drug Recognition Expert (DRE) Validation Study, Eugene V. Adler and Marcelline Burns (Arizona - 1994)

6. Laboratory Validation Study of Drug Evaluation and Classification Program, Ethanol, Cocaine, and Marijuana, Heishman SJ, Singleton EG, Crouch DJ (1996)

7. Laboratory Validation Study of Drug Evaluation and Classification Program: Alprazolam, d-Amphetamine, Codeine, and Marijuana, Stephen J. Heishman, Edward G. Singleton, Dennis J. Crouch (1998)