FMCSA Random Drug
Testing Compliance

Who belongs in the random pool

The rule on this point is short and strict: covered employees, and only covered employees, go in the pool (§ 382.305(j)(1)). In practice that means every driver who holds a CDL and drives a commercial motor vehicle for you. FMCSA’s own guidance says part-time CDL drivers stay in the pool at all times, and a driver who works for two carriers belongs in both carriers’ pools.

Drivers who don’t need a CDL for the vehicle they operate stay out of the DOT pool. You can test them under company policy if you like, but that testing has to run in a completely separate pool and can’t be presented as DOT testing.

Owner-operators

This is where small operations get caught. A company with more than one CDL driver can manage its own random program. An owner-operator with a single CDL driver, who is not leased to another motor carrier, has to place that driver in a consortium. The Clearinghouse registration process makes the same point: owner-operators must designate a consortium or third-party administrator (C/TPA).

The 2026 rates and when they change

For 2026 the minimum annual rates are 50 percent for controlled substances and 10 percent for alcohol, each measured against the average number of driver positions you had during the year (§ 382.305(b)). DOT’s random testing rate table lists both as unchanged since 2020, so no new FMCSA notice was needed for this year. If you have an older checklist that shows a 25 percent controlled-substances rate, it’s out of date; the rate was lower in earlier years and went back to 50 percent in 2020.

The figures aren’t fixed. FMCSA reviews the industry’s random positive and violation rates from the annual MIS reports and publishes any change in the Federal Register, effective January 1 of the following year.

Because a change takes effect on January 1, the useful habit is to check DOT’s rate table each December and January instead of carrying last year’s number forward.

Turning a rate into a test count

Your annual target comes from the average size of your pool, not the headcount on January 1. Add up the number of eligible drivers for each testing period and divide by the number of periods. Employers who select more often than monthly only need to work this out once a month (§ 382.305(j)(1)).

A fleet that averages 40 driver positions across the year needs at least 20 random controlled-substance tests and 4 random alcohol tests. A fleet that grows from 30 drivers to 50 over the year averages about 40 and plans on the same numbers, not on its year-end headcount. Only random tests count toward the target; post-accident and reasonable-suspicion tests don’t.

Using a consortium or service agent

A service agent such as a C/TPA can run the draws for you, and your drivers can sit in a larger pool alongside other employers’ drivers (§ 382.305(j)(2)). You still carry the responsibility. You need to make sure the agent tests at the required rate and that only covered drivers are in the pool.

How a selection should run

Selection has to use a scientifically valid method, such as a random number table or a computer random number generator matched to identifiers like payroll numbers. Every driver must have an equal chance of being picked each time you draw (§ 382.305(i)). Being picked last month doesn’t take anyone out of this month’s draw. A driver selected in March goes back into the pool and can be selected again in April.

That rules out several things carriers do with good intentions: rotating through the roster, skipping people who tested recently, or letting a dispatcher pick someone who “is around today.” If you draw by terminal or location, FMCSA expects a two-step process. Select the locations first, then select employees at those locations, and make sure every driver in the pool still had an equal chance regardless of where he or she works.

Timing and notification

Tests are unannounced and must be spread reasonably across the calendar year (§ 382.305(k)). Many employers draw monthly or quarterly so they never face a December scramble. Once notified, the driver must go to the collection site immediately. A driver who is doing a safety-sensitive job other than driving, such as loading, has to stop and go as soon as possible (§ 382.305(l)).

A driver selected in one selection period has to be tested in that period. Testing in a later period is a violation, though the test still counts toward the annual minimum. An alternate can be used only if the selected driver will be unavailable for the whole period, for example because of a layoff, illness, injury or vacation.

Records and the Clearinghouse

Section 382.401 sets three retention tiers that matter for random testing.

The regulation doesn’t spell out exactly what “documents relating to the random selection process” means, so be generous. For every draw, a reviewer should be able to see who was in the pool, who was selected, when the draw happened, how it was made, and when each selected driver was notified and tested.

What reaches the Clearinghouse

A verified positive from a random test is reported by the MRO, within two business days of verification. You report refusals, alcohol results of 0.04 or higher, and actual knowledge of a violation, by the close of the third business day after you learn of it. A driver with an unresolved violation cannot perform safety-sensitive functions. You also have to query each current CDL driver at least once a year, with the driver’s consent, and run a full query before you put a new driver to work.

One detail that confuses year-end reporting: a test counts toward the MIS year in which the specimen was collected, not the year the MRO verified the result.

Where programs go wrong

• The pool includes people who aren’t covered, such as office staff or non-CDL drivers, which breaks the “only covered employees” rule and dilutes the real drivers’ chance of selection.
• The roster lags behind hiring and terminations, so new drivers aren’t in the pool for weeks and former drivers still get selected.
• Selected drivers are tested after the selection period ends because the draw and the notification weren’t coordinated.
• Someone with authority picks the “random” driver by hand.
• An alcohol test is given to a driver who is off duty.
• The draw happened but nobody saved the roster it was drawn from, so there is nothing to show an auditor.
• A single-driver owner-operator runs a personal “program” and never joins a consortium.

If your drivers also fall under another DOT agency

A driver covered by more than one DOT agency is tested at the rate set by the agency that regulates more than half of the driver’s safety-sensitive work. An employer in that position can keep separate pools for each required rate, or put everyone in one pool and test at the highest applicable rate (§ 382.305(n) and (o)). The most common overlap is a pipeline operator with CDL drivers; see the PHMSA page for its drug-only rule. Transit employers have their own rules on the FTA page. For a side-by-side view of every agency’s 2026 rates, use the comparison chart.