A controlled human study found no statistically significant difference in overall simulated driving performance after a single oral dose of d-methamphetamine compared with placebo. That finding does not establish that methamphetamine is safe for driving. It answers a narrower question about a small group, one dose, a short simulator assessment and a particular testing window.
Silber and colleagues’ 2012 study is useful in forensic toxicology because it reports both a nonsignificant overall result and differences in individual driving measures. Reading those results together avoids overstating either impairment or the absence of impairment.
Twenty adults with previous recreational stimulant exposure—10 men and 10 women, aged 21–32—completed two sessions. Each received oral d-methamphetamine at 0.42 mg/kg in one session and matching placebo in the other. The experiment used a double-blind, counterbalanced, repeated-measures design with a two-week washout. Participants served as their own comparison.
The simulator assessment, described as occurring approximately 2.5 hours after dosing, involved freeway and city scenarios in daytime and nighttime conditions. Four tasks lasted approximately five minutes each. The researchers combined weighted error measures into an overall driving score and also examined 35 individual variables.
| Study feature | Why it matters |
|---|---|
| 20 participants with prior stimulant experience | The sample does not represent every driver or every pattern of methamphetamine use. |
| Single oral dose of 0.42 mg/kg | The study did not reproduce repeated dosing or other routes of administration. |
| Approximately 20 minutes of simulated driving | The assessment differs from prolonged driving on real roads. |
| Placebo-controlled crossover comparison | Comparing each participant across conditions reduces some between-person differences. |
The overall weighted driving-error score averaged 102.4 after d-methamphetamine and 95.3 after placebo. The reported comparison was not statistically significant (p = 0.19). The study therefore did not demonstrate a significant overall performance decrement under its experimental conditions.
A nonsignificant comparison is not an equivalence test. The result does not prove that the two conditions produced identical performance, and it does not rule out effects too small for this experiment to detect. The authors identified the small sample and possible limits of the simulator’s demands as reasons for caution.
Participants drove significantly more slowly on the freeway when an emergency situation occurred after d-methamphetamine than after placebo (p < 0.05). Average speeds in both conditions nevertheless exceeded the 100-km/h limit. A difference in this one measure should not be relabeled as an overall improvement in driving safety.
The methamphetamine condition also produced approximately four times as many failures to stop at red lights. That comparison was not statistically significant (p = 0.11). The numerical difference is relevant to describing the findings, but it does not establish a proven drug effect on red-light compliance. The study examined 35 individual variables, which also calls for caution when interpreting an isolated significant comparison.
| Reported finding | Supported interpretation |
|---|---|
| Overall driving score: p = 0.19 | No statistically significant overall difference was demonstrated. |
| Lower freeway speed at the emergency event: p < 0.05 | A specific driving measure differed; this does not establish global safety. |
| More red-light failures: p = 0.11 | A numerical increase occurred without a statistically significant comparison. |
At 120 minutes after administration, mean methamphetamine concentrations were 72 ng/mL in blood and 285 ng/mL in saliva. At 170 minutes, the respective means were 67 and 223 ng/mL; at 240 minutes, they were 59 and 190 ng/mL. These are group means in different biological matrices, not interchangeable values or individual impairment thresholds.
The paper describes gas chromatography–mass spectrometry for specific amphetamine measurements. It also contains differences between its experimental-procedure and sampling-section descriptions of collection times. The numerical results should therefore be tied to their explicitly reported post-dose times, without inventing a more precise sampling chronology.
This experiment provides evidence against an absolute claim that every methamphetamine exposure necessarily produces a detectable decrement on every driving measure. It also provides no basis for declaring a particular driver unimpaired merely because a blood result resembles the study’s group mean.
A case-specific assessment must address the actual driving evidence, timing, specimen, analytical result and circumstances. This experiment cannot establish what an individual’s concentration was earlier, whether an observed error was caused by methamphetamine, or how repeated dosing, sleep loss or additional substances would change performance.
For the related question of improvements on selected laboratory tasks, see Can Methamphetamine Improve Performance? What Controlled Human Studies Show. Laboratory-task benefits and overall driving ability are different outcomes.
No. It found no statistically significant overall difference in a specific small simulator experiment. It did not establish road safety.
No. The reported comparison was p = 0.11. The numerical increase should be reported with that limitation.
No. Group concentrations from this experiment do not establish a diagnostic threshold for individual drivers.
No. The administered drug was d-methamphetamine. Findings should retain that drug and formulation rather than being generalized to every amphetamine preparation.
Silber BY, Croft RJ, Downey LA, Papafotiou K, Camfield DA, Stough C. The effect of d-methamphetamine on simulated driving performance. Human Psychopharmacology: Clinical and Experimental. 2012;27:139–144. doi:10.1002/hup.1238.