Drug and Alcohol Testing: What the Programme Actually Requires
Not having a programme at all is one of the violations that fails a safety audit outright. Setting one up is neither difficult nor expensive — but a few details decide whether yours would survive inspection.
The random testing rates
49 CFR § 382.305 sets minimum annual random testing rates as a percentage of average driver positions:
| Test | Baseline minimum | Range the Administrator may set |
|---|---|---|
| Controlled substances | 50% | 25% or 50%, based on the industry positive rate |
| Alcohol | 10% | 10%, 25% or 50%, based on the industry violation rate |
The rates are not fixed forever. FMCSA sets them from data reported by employers, and when they change, the new minimum annual percentage rate is published in the Federal Register and takes effect on 1 January of the following year. For 2026 the FMCSA rates are unchanged at 50 percent for drugs and 10 percent for alcohol. Check the current rate rather than assuming last year's.
The steps are set by thresholds. The drug rate drops from 50 to 25 percent only after two consecutive years with an industry positive rate below 1.0 percent, and goes back to 50 percent in any year it reaches 1.0 percent. The alcohol rate moves between 10, 25 and 50 percent on violation rates of 0.5 and 1.0 percent.
Working out how many tests
The rate applies to the average number of driver positions, not to the drivers on the books on 31 December. FMCSA's guidance gives the formula: add up the drivers subject to testing in each selection period (D), divide by the number of periods in a full year (P), and multiply by the rate.
FMCSA's own example: a carrier starts operating in August and draws quarterly. It has 0 drivers in the first two quarters, 16 in the third and 12 in the fourth. D is 28, P is 4, so the average is 7, and 50 percent of 7 is 3.5 — rounded up to 4 drug tests, spread as two in each of the last two quarters. The alcohol requirement is 10 percent of 7, or 0.7, which rounds up to one test.
In a consortium the arithmetic is done on the whole pool, which is why an individual owner-operator might be selected twice in a year or not at all.
What “random” has to mean
The regulation is specific, and these are the details that fail an audit:
- Selections must be unannounced.
- The method must be scientifically valid, such as a random number table or computer generator matched to payroll or similar numbers, giving every driver an equal chance of selection each time a draw is made.
- Every driver selected must be tested in that selection period.
- Testing dates must be spread reasonably throughout the calendar year.
- A selected driver goes to the collection site immediately once notified.
- Random alcohol tests may only be done just before, during or just after the driver performs safety-sensitive functions.
The spreading requirement is the trap. A carrier who runs all its random selections in November has met the percentage and failed the requirement. The regulation does not name a frequency, but quarterly draws are the common way to show the dates are spread, and an auditor can see your dates.
And “equal chance each time” means a driver selected in March goes back in the pool for the next draw. Removing already-tested drivers feels fair and breaks the rule.
How a one-truck carrier complies
You cannot run a random selection pool of one person in any meaningful way, which is why owner-operators join a consortium or third-party administrator. The consortium pools you with other drivers, runs the selections, arranges collections and keeps the records. The regulation allows this as long as the pool as a whole is tested at the required rate.
This is also how an owner-operator meets the Clearinghouse query obligations, since you cannot act as your own employer for that purpose.
What you must keep: the enrolment documentation, the selection and test records, and evidence the programme actually ran. Enrolling and never being drawn is fine; enrolling and having no records is not. You also need a written policy and educational materials under § 382.601, covering who is covered, prohibited conduct, when tests happen, what counts as refusal and the consequences, with a signed receipt from each driver — yourself included.
When testing happens besides random
| Test | When | Key detail |
|---|---|---|
| Pre-employment | Before the driver first performs safety-sensitive functions | Drug test with a verified negative result required; pre-employment alcohol testing is optional but, if used, must apply to everyone |
| Post-accident | After qualifying crashes (below) | Alcohol within 2 hours if possible, stop trying at 8; drugs within 32 hours |
| Reasonable suspicion | Based on a trained supervisor's specific observations | Supervisors need at least 60 minutes of alcohol training and 60 minutes on controlled substances |
| Return-to-duty and follow-up | After a violation and completion of the substance abuse professional process | Follow-up testing as directed by the substance abuse professional |
The pre-employment drug test can be skipped only if the driver was in a compliant testing programme within the previous 30 days, was tested within the past 6 months or was in a random pool for the previous 12 months, and no prior employer has a violation on record in the last six months. You must contact the programme and document all of that, so for most small carriers simply testing is easier.
Post-accident: when a test is required
| What happened | Citation for a moving violation? | Test required |
|---|---|---|
| A fatality | Not relevant | Yes |
| Injury treated away from the scene | Yes | Yes |
| Injury treated away from the scene | No | No |
| A vehicle towed away with disabling damage | Yes | Yes |
| A vehicle towed away with disabling damage | No | No |
If the alcohol test is not done within 2 hours, or either test is abandoned at the time limit, write down why and keep the record. Breath, blood or urine tests done by police with independent authority can count, if they meet the requirements and you obtain the results. See also the accident register.
Results below the violation line
An alcohol result of 0.02 or more but below 0.04 is not a violation, but the driver must be removed from safety-sensitive functions for at least 24 hours after the test, or until the next regularly scheduled duty period if that is later.
How long to keep the records
| Record | Keep for |
|---|---|
| Alcohol results of 0.02 or more, verified positive drug results, refusals, evaluations and referrals, annual summaries | 5 years |
| Collection process records | 2 years |
| Negative and cancelled drug results, alcohol results below 0.02 | 1 year |
| Supervisor and driver training records | While the person performs the function, plus 2 years |
Records must be kept securely and produced at your principal place of business within two business days of an FMCSA request.
Where this shows up
Failing to implement an alcohol and controlled substances testing programme, or a random testing programme, is on the list of violations that cause automatic failure of a new entrant safety audit. So is using a driver known to have tested positive, refused a test, or had an alcohol concentration of 0.04 or more. Each is a single-occurrence failure, whatever else your operation looks like.
It also feeds the Controlled Substances and Alcohol BASIC, which is one of the five publicly visible categories.
One thing worth knowing if you are below the CDL threshold: Part 382 testing applies to drivers required to hold a CDL. A box truck under 26,001 pounds generally falls outside it — though customers and insurers often require a programme by contract anyway.
Common mistakes
- Enrolling in a consortium after the first load instead of before, with no pre-employment result on file.
- No signed policy receipt for the owner-operator.
- Sending the driver for testing “later today” after notification of a random selection.
- No written reason when a post-accident test could not be done in time.
Frequently asked questions
What are the random testing rates?
The regulation sets a baseline of 50 percent of average driver positions for controlled substances and 10 percent for alcohol, with the Administrator able to adjust within defined ranges. For 2026 FMCSA kept them at 50 and 10 percent. Changes are published in the Federal Register and take effect on 1 January.
Can I do all my random tests at once?
No. Testing dates must be spread reasonably throughout the calendar year.
Does a driver already tested come out of the pool?
No. Every driver must have an equal chance of selection each time a draw is made.
How does a single owner-operator comply?
By joining a consortium or third-party administrator that operates the random pool and keeps the records, and by keeping a written policy with a signed receipt.
When is a post-accident test required?
After any crash involving a fatality, and after a crash with an injury treated away from the scene or disabling tow-away damage if the driver receives a citation for a moving violation.
Related
Watch
Return-to-Duty Process for CDL Drivers — FMCSA. The video loads from YouTube only when you press play.
Sources
- 49 CFR § 382.305 — Random testing.
- 49 CFR Part 385 Subpart D — eCFR.
- 2026 DOT random testing rates — U.S. DOT Office of Drug and Alcohol Policy and Compliance.
- Random testing for a partial calendar year — FMCSA.
- 49 CFR § 382.301 — Pre-employment testing — eCFR.
- 49 CFR § 382.303 — Post-accident testing — eCFR.
- 49 CFR § 382.401 — Retention of records — eCFR.
- 49 CFR § 382.505 — Other alcohol-related conduct — eCFR.
- 49 CFR § 382.601 — Employer obligation to promulgate a policy — eCFR.
- 49 CFR § 382.603 — Training for supervisors — eCFR.
- 49 CFR § 385.321 — Failure of safety audit — eCFR.
General information, not legal advice. Regulations change; confirm current requirements with the agency before acting.