DOT Random Testing Explained

The numbers for 2026

If you run a fleet, you've probably heard some version of "you have to randomly test your drivers." Fewer people can tell you exactly what that means in practice — how the number gets calculated, who has to run the draw, or what an auditor actually checks.

The Federal Motor Carrier Safety Administration sets random testing rates every year based on industry-wide positive test data. For 2026, the rates didn't move: 50% DRUGS / 10% ALCOHOL of your average driver pool has to be tested over the course of the year. That's the sixth year in a row at these levels — FMCSA raised the drug rate from 25% to 50% back in January 2020, and it's stayed there since.

The rule that keeps it there is simple. If the industry's positive rate stays at or above 1%, the rate holds at 50%. Bring it below 1% for two straight years, and FMCSA can lower it back down. That hasn't happened, and marijuana alone still accounts for roughly 60% of all positive results going back to 2020.

What "50%" actually means for your pool

This is where a lot of small fleets get tripped up. The 50% isn't "half your drivers get tested once." It's calculated against your average number of driver positions across the year, and the draws happen throughout the year — not all at once, and not on a schedule anyone can predict. A driver could be selected more than once in the same year. That's the point: the selection has to be random enough that no one, including you, can game it.

The regulation 49 CFR 382.305 requires a "scientifically valid method" for the draw — meaning every driver in your pool has an equal chance each time, not a rotation, not a list, and not someone's judgment call.

Why the gap between passing and using matters

Here's a number that gets less attention than it should: in 2024, fentanyl showed up more than seven times as often in random tests as in pre-employment tests. That gap is exactly what random testing exists to catch. A driver who's clean at hiring isn't necessarily clean six months later, and pre-employment screening alone won't tell you that.

What an audit actually looks for

When FMCSA or a state inspector reviews your program, they're not just checking that testing happened. They're checking that:

• Your selection method meets the "scientifically valid" standard, not just "we tried to be fair"
• You can produce records showing who was in the pool, who was drawn, and when
• Your rate for the year actually met the 50%/10% minimums — not an estimate, an actual count

If your program runs on a spreadsheet someone updates manually, this is usually where things fall apart. Not because the testing itself was wrong, but because the paper trail doesn't hold up to the question "prove it."

Where this leaves small fleets

If you're running under 20 drivers, the math still applies to you the same way it applies to a 200-truck operation — you don't get a pass on the percentage just because your pool is small. What changes is how much sense it makes to pay a TPA's flat monthly fee versus a system that scales with your actual driver count. That's a separate conversation from the compliance requirement itself, and worth having with clear eyes about what you're actually paying for.