A survey of 26,458 US workers found 7.4% used cannabis at or within 2 hours of work, with higher rates in recreational-legal states (8.5%), high-risk jobs (11.4%), and among medical cardholders (39%).
Employers, occupational safety researchers, and policymakers.
39% of workers with medical authorization used cannabis at or near work
What the researchers found
Workplace use highest in recreational states (8.5%) vs. medical-only (6.3%) or illegal (6.2%). High-risk workers: 11.4% vs. 5.8%. Medical users: 29.4% vs. recreational 15.6%. Medical authorization: 39.0% vs. 17.4% without.
Why it matters
Quantifies who uses cannabis at work as legalization expands. High-risk workers and medically authorized users have the highest rates.
The numbers in context
26,458 workers; 7.4% overall; 11.4% high-risk jobs; 39.0% with medical authorization; 8.5% in recreational states.
How the study worked
Cross-sectional data from ICPS wave 6 (2023). 26,458 respondents aged 16-65. Regression models by state laws, occupational risk, and medical authorization.
What this study cannot tell us
Cross-sectional. Self-reported. "Within 2 hours" is broad. Impairment not measured.
How to read the evidence
Large nationally representative sample with robust analysis.
When this study was published
2025 publication using 2023 ICPS data.
The bigger picture
Medical users show the highest workplace rates, suggesting accommodation policies must balance access with safety.
Questions still open
- Does workplace use translate to accidents?
- How to balance medical accommodation with safety?
Common questions
How common is cannabis use at work?
Do high-risk workers use more at work?
Read the original research
Cross-sectional analysis of cannabis use at work in the USA: differences by occupational risk level and state-level cannabis laws.
BMJ public health, 3(2), e001589
Citation
Kucera, Ava; Hammond, David. (2025). Cross-sectional analysis of cannabis use at work in the USA: differences by occupational risk level and state-level cannabis laws.. BMJ public health, 3(2), e001589. https://doi.org/10.1136/bmjph-2024-001589