FMCSA sets minimum annual rates for random drug and alcohol testing of CDL drivers. As of 2026, the minimums are 50% for controlled substances and 10% for alcohol, set by 49 CFR §382.305.
But what does “50%” mean? It’s not “half your drivers get tested.” Here’s how the math works.
The percentages apply to driver positions, not drivers
The rate is measured against the average number of driver positions over the year. Selections are drawn randomly, with replacement. A five-driver fleet at the 50% drug rate generates roughly two to three selections per year. Because every draw is random, the same driver can be selected twice while another isn’t selected at all.
That’s by design. A driver who just tested is immediately back in the pool. Nobody is ever “safe until next year.”
Selections must be spread through the year
Random testing only works if it’s unpredictable. FMCSA requires selections to be reasonably spread across the calendar. A consortium typically draws quarterly. Every selection gets documented, and so does every reason a selected driver couldn’t test. Auditors ask for both.
What the rates mean for a one-truck operation
An owner-operator can’t run a random pool of one. FMCSA requires selections from a pool of two or more covered drivers (49 CFR §382.103(b)). In a consortium pool, the rates apply across the whole shared pool. Some years you’ll be selected, some years you won’t. Both outcomes are compliant, because the pool as a whole hits the minimums.
The rates can change — here’s the mechanism
FMCSA adjusts the minimums based on industry-wide violation rates reported each year. The drug rate drops to 25% only after the industry positive rate stays under 1% for two straight years. It snaps back to 50% when the rate crosses 1%. That’s exactly what happened for 2020, after the 2018 survey. Any change is published in the Federal Register and takes effect January 1.
Current minimums (2026): 50% drugs / 10% alcohol. We update this post when FMCSA changes the rates.
What you actually have to do
If you’re in a properly run consortium: nothing. Selections happen at the required rates, you get a call when a driver is drawn, and the documentation lands in your audit file. If you’re self-administering, you own the draw methodology, the spread, and the paper trail. A defensible version of that is real work.
Not sure your current program hits the minimums? Ask us to look — (651) 482-1124.