Document Type


Publication Date



This article is the author's final published version in JAMA Psychiatry, Volume 80, Issue 9, September 2023, Pages 914 - 923.

The published version is available at

Copyright © 2023 Marcotte TD et al. JAMA Psychiatry.

This is an open access article distributed under the terms of the CC-BY License.


IMPORTANCE: With increasing medicinal and recreational cannabis legalization, there is a public health need for effective and unbiased evaluations for determining whether a driver is impaired due to Δ9-tetrahydrocannabinol (THC) exposure. Field sobriety tests (FSTs) are a key component of the gold standard law enforcement officer-based evaluations, yet controlled studies are inconclusive regarding their efficacy in detecting whether a person is under the influence of THC.

OBJECTIVE: To examine the classification accuracy of FSTs with respect to cannabis exposure and driving impairment (as determined via a driving simulation).

DESIGN, SETTING, AND PARTICIPANTS: This double-blind, placebo-controlled parallel randomized clinical trial was conducted from February 2017 to June 2019 at the Center for Medicinal Cannabis Research, University of California, San Diego. Participants were aged 21 to 55 years and had used cannabis in the past month. Data were analyzed from August 2021 to April 2023.

INTERVENTION: Participants were randomized 1:1:1 to placebo (0.02% THC), 5.9% THC cannabis, or 13.4% THC cannabis smoked ad libitum.

MAIN OUTCOME AND MEASURES: The primary end point was law enforcement officer determination of FST impairment at 4 time points after smoking. Additional measures included officer estimation as to whether participants were in the THC or placebo group as well as driving simulator data. Officers did not observe driving performance.

RESULTS: The study included 184 participants (117 [63.6%] male; mean [SD] age, 30 [8.3] years) who had used cannabis a mean (SD) of 16.7 (9.8) days in the past 30 days; 121 received THC and 63, placebo. Officers classified 98 participants (81.0%) in the THC group and 31 (49.2%) in the placebo group as FST impaired (difference, 31.8 percentage points; 95% CI, 16.4-47.2 percentage points; P < .001) at 70 minutes after smoking. The THC group performed significantly worse than the placebo group on 8 of 27 individual FST components (29.6%) and all FST summary scores. However, the placebo group did not complete a median of 8 (IQR, 5-11) FST components as instructed. Of 128 participants classified as FST impaired, officers suspected 127 (99.2%) as having received THC. Driving simulator performance was significantly associated with results of select FSTs (eg, ≥2 clues on One Leg Stand was associated with impairment on the simulator: odds ratio, 3.09; 95% CI, 1.63-5.88; P < .001).

CONCLUSIONS AND RELEVANCE: This randomized clinical trial found that when administered by highly trained officers, FSTs differentiated between individuals receiving THC vs placebo and driving abilities were associated with results of some FSTs. However, the high rate at which the participants receiving placebo failed to adequately perform FSTs and the high frequency that poor FST performance was suspected to be due to THC-related impairment suggest that FSTs, absent other indicators, may be insufficient to denote THC-specific impairment in drivers.

TRIAL REGISTRATION: Identifier: NCT02849587.