rethinkTHC Search
Menu
Study breakdown

Even Multiple Blood Draws Cannot Reliably Determine When Someone Last Used Cannabis

Retrospective CohortModerate evidence
The takeaway

Analysis of 35 impaired driving cases with multiple blood draws showed that THC concentrations sometimes increased at later time points, making it difficult to determine when cannabis was last used.

Forensic toxicologists, defense attorneys in cannabis-DUI cases, policymakers setting THC driving limits

13.6% of samples showed THC increases at later time points

What the researchers found

In 35 DUID cases with multiple blood draws (81 total samples), THC concentrations ranged from 0.74 to 40 ng/mL. Notably, 11 samples showed an increase in THC concentration at a later collection time point, which undermines the assumption that THC levels consistently decline after use. Collection times ranged from 32 minutes to over 12 hours after the incident.

Why it matters

Many jurisdictions set per se THC limits for driving, similar to blood alcohol limits. This study demonstrates a fundamental problem with that approach: unlike alcohol, THC blood levels do not reliably decline in a predictable pattern, making it impossible to back-calculate the level at the time of driving.

The numbers in context

35 cases, 81 samples. THC range: 0.74-40 ng/mL. 11-OH-THC range: 1.0-16 ng/mL (60 samples). THC-COOH range: 7.1-470 ng/mL. Collection times: 00:32-12:42 hours post-incident. 11 samples (13.6%) showed THC increase at a later time point.

How the study worked

Retrospective analysis of cannabinoid-positive driving under the influence cases from 2019-2023 that had multiple blood draws. Samples were analyzed using LC-MS/MS with reporting limits of 0.5 ng/mL for THC, 1.0 ng/mL for 11-OH-THC, and 5.0 ng/mL for THC-COOH.

What this study cannot tell us

Relatively small sample (35 cases). Cannot control for subjects' cannabis use between blood draws in all cases. No information on the specific cannabis products used, route of administration, or chronic use history. Forensic casework data with inherent variability in collection timing and conditions.

How to read the evidence

Moderate evidence from real-world forensic casework data, though limited by small sample size and uncontrolled variables.

When this study was published

2025 study analyzing DUID cases from 2019-2023.

The bigger picture

As more jurisdictions implement cannabis-impaired driving laws, the scientific limitations of blood THC testing become increasingly important. This study provides real-world DUID casework evidence that even the gold-standard approach of multiple blood draws cannot reliably establish the time of cannabis use, challenging the validity of per se THC limits.

Questions still open

  • What causes THC levels to increase at later time points? Is redistribution from fat stores responsible? Should per se THC driving limits be abandoned in favor of impairment-based assessments? How should forensic toxicologists interpret rising THC levels in casework?

Common questions

Can a blood test tell when someone last used cannabis?
This study shows that even with multiple blood draws from the same person, it is often impossible to determine when cannabis was last used. In about 14% of samples, THC levels actually increased at later time points, contradicting the assumption that levels consistently decline after use.
Why do THC levels sometimes go up?
THC is stored in fat tissue and can be re-released into the blood later. This redistribution effect, combined with individual variation in metabolism, means that a single or even multiple blood THC levels cannot reliably indicate when cannabis was consumed.

Read the original research

Changes in blood cannabinoid concentrations over multiple collection times in driving under the influence of drugs casework.

Journal of analytical toxicology, 49(8), 576-586

Citation

Peterson, Brianna L; Hessler, Meaghan R. (2025). Changes in blood cannabinoid concentrations over multiple collection times in driving under the influence of drugs casework.. Journal of analytical toxicology, 49(8), 576-586. https://doi.org/10.1093/jat/bkaf052

Explore the wider topic