Analysis of 740 cannabis-involved driving cases in Germany found a weak relationship between THC blood levels and impairment, with no useful cut-off value identifiable, and the new 3.5 ng/mL threshold would eliminate about one-third of current offenses.
Policymakers setting cannabis driving limits and drivers who use cannabis.
No useful cut-off value could be defined due to too much overlap between impaired and non-impaired drivers' THC levels
What the researchers found
THC serum concentrations were slightly but significantly higher in cases with criminal driving offenses compared to those without, but the absolute differences were small with too much overlap between groups to define a useful cut-off. Under Germany's new 3.5 ng/mL threshold (raised from 1 ng/mL), approximately one-third of evaluated cases would no longer be subject to legal penalties.
Why it matters
Germany recently raised its legal THC driving threshold from 1 to 3.5 ng/mL. This study examines whether any specific blood THC level reliably indicates impairment, with implications for driving laws worldwide.
The numbers in context
740 cases analyzed. Slightly higher THC and CIF values in cases with criminal offenses (statistically significant but small absolute differences). Too much overlap between impaired and non-impaired groups for a useful cut-off. ~33% of cases would no longer face penalties under the new 3.5 ng/mL threshold.
How the study worked
Retrospective analysis of 740 cases from 2020-2021 where blood samples were submitted for toxicological analysis from drivers suspected of driving under the influence of cannabis alone (DUIC) or combined with alcohol (DUIAC). Evaluated behavioral and driving impairments against THC serum concentrations and cannabis influence factor (CIF) values.
What this study cannot tell us
Retrospective design relies on police assessment of impairment, which is subjective. Only cases where police suspected cannabis involvement were included, creating selection bias. 2020-2021 data may not represent current patterns.
How to read the evidence
Moderate: large forensic dataset (740 cases) with objective blood measurements, though limited by retrospective design and subjective impairment assessment.
When this study was published
2025 study using 2020-2021 data.
The bigger picture
This adds to international evidence that blood THC levels are a poor proxy for driving impairment. Unlike blood alcohol concentration, which has a more predictable relationship with impairment, THC levels do not translate reliably to functional impairment.
Questions still open
- Should cannabis driving laws abandon per se THC limits in favor of behavioral impairment testing? Is the new 3.5 ng/mL threshold better justified than the old 1 ng/mL? Could standardized impairment tests replace or supplement blood testing?
Common questions
Can a blood test tell if someone is too impaired to drive from cannabis?
What changed with Germany's new cannabis driving law?
Read the original research
Cannabis in road traffic - a retrospective analysis to identify possible cut-off-values.
Forensic science, medicine, and pathology
Citation
Ludwig, A; Küpper, U; Lau, M; Holzer, A; Wulff, T; Hartung, B. (2025). Cannabis in road traffic - a retrospective analysis to identify possible cut-off-values.. Forensic science, medicine, and pathology. https://doi.org/10.1007/s12024-025-01133-1