When a worker shows up through a staffing agency instead of a direct hire, drug testing responsibility gets murky fast. Both the agency and the company using the worker often assume the other one is handling it. That gap is where programs actually break down, not because anyone intended to skip testing, but because nobody wrote down who owns which piece.
This matters more as staffing relationships become a normal part of how companies fill roles, from short-term warehouse help to specialized technical placements. A testing program that works fine for direct employees can quietly stop functioning the moment a meaningful share of the workforce arrives through a third party.
Two different frameworks, two different answers
How this question gets answered depends heavily on whether the role is subject to DOT testing rules or not.
For commercial driving positions that require a CDL, FMCSA has specifically addressed staffing arrangements. Federal guidance states plainly that a driver staffing agency supplying commercial drivers to operate a CMV can itself qualify as an “employer” under the federal motor carrier safety regulations, which means it can carry direct drug and alcohol testing obligations rather than simply passing the responsibility along to the motor carrier that ultimately uses the driver. That guidance exists because enough confusion had built up in the industry about who was actually on the hook when a driver came from a staffing firm rather than a traditional hire (FMCSA, Controlled Substances and Alcohol Testing Responsibilities of Commercial Driver Staffing Agencies and Motor Carriers That Use Them).
In practice, this means a motor carrier using CDL drivers supplied by a staffing agency cannot simply assume the agency has it covered, and the agency cannot assume the carrier is testing on its end. Both parties need a clear, written understanding of exactly which entity is running the random testing pool the driver is enrolled in, who is responsible for pre-employment testing before the driver starts work, and who handles post-accident and reasonable suspicion testing if an incident happens on-site. Enrolling the same driver in two separate, uncoordinated random pools, or in none at all because each side assumed the other handled it, are both real failure modes that show up when this isn’t addressed directly.
The non-DOT situation is less defined, but still needs an answer
Outside of DOT-regulated roles, there’s no federal rule spelling out who is responsible for testing a staffing agency’s placed worker. That absence of a clear rule is exactly why so many companies get this wrong. Without a federal default to fall back on, the responsibility has to be assigned explicitly in the business relationship between the agency and the client company, or it won’t get assigned at all.
In most general-industry staffing arrangements, one of two models is in play. In the first, the staffing agency screens candidates before ever placing them, and the client company agrees to accept that screening in place of running its own. In the second, the client company treats the placed worker like any other person stepping onto the worksite and runs its own program-level testing regardless of what the agency already did. Neither model is inherently wrong, but a program only works when everyone involved, including the worker, understands which model is actually being used.
What should be in writing before the first placement happens
A handful of specific points should be settled and documented before an agency sends its first worker to a client site, not worked out after a problem comes up.
The agreement should state plainly which party is responsible for pre-placement testing, and whether the client company will accept the agency’s test result or require its own. It should also state what happens for testing situations that come up after a worker is already on-site, such as random selection, reasonable suspicion, and post-accident testing. A worker who is on an agency’s payroll but working daily under a client company’s direct supervision creates a genuine question about who is positioned to observe behavior and make a reasonable suspicion call in the first place, and that needs an answer before anyone is relying on it in the moment.
Specimen type and turnaround expectations are also worth settling up front, especially for short-term or high-volume placements where a multi-day lab turnaround can functionally defeat the purpose of pre-placement testing. A collection method that produces a result quickly fits naturally into a staffing workflow where a candidate may need to start work within a day or two of being screened.
Finally, the agreement should specify what documentation the agency will share with the client company, and what stays with the agency. A client company that later needs to show a regulator or an insurer that a worker was properly screened cannot do that if the only record of the test lives in a staffing agency’s file that it never sees.
Seasonal and high-volume placements add another layer
The ambiguity gets worse, not better, when staffing is used for seasonal surges or short-term projects rather than a single ongoing placement. A warehouse bringing on forty temporary workers for a peak season, or a construction site rotating through subcontracted labor on a short job, creates a lot more opportunities for a testing gap to go unnoticed simply because of volume and turnover. When a handful of individual placements become dozens happening at once, an informal understanding between an account manager and a site supervisor stops being a workable substitute for a written process.
This is also where turnaround time becomes a practical, not just theoretical, concern. A testing process that takes several days to return a result can mean workers are already on-site and working before their screening is even back, which defeats the purpose of pre-placement testing in the first place. Matching the collection method and lab process to the actual pace of the placement, rather than defaulting to whatever process the agency happens to already use for its slower-moving direct placements, is worth revisiting whenever the volume or timeline changes.
Common mistakes
The most common mistake is simply never discussing it, with each side defaulting to an assumption that favors less work for itself. A close second is a written agreement that says testing “will be coordinated” without naming who does what, which functions the same as having no agreement at all once an actual incident happens. Another frequent problem is treating a one-time placement conversation as permanent, when staffing relationships change vendors, account managers, and internal practices more often than direct-hire programs do. What was true about an agency’s testing practices two years ago may not be true today.
A narrower but costly mistake, specific to DOT-regulated roles, is a motor carrier assuming that because a driver arrived through a reputable staffing agency, the agency has necessarily enrolled that driver in a FMCSA-compliant random testing program. That assumption needs to be verified, not inferred from the agency’s general reputation.
A common misconception
A common misconception is that using a staffing agency somehow transfers all drug testing liability away from the client company. It doesn’t work that way in practice. Regulators and courts generally look at who actually controls the work and the worksite, not just who issues the paycheck, when sorting out where responsibility sits. A client company that never checks whether its staffing partner is actually testing, and never puts anything in writing about it, is still exposed if something goes wrong on its site.
Keeping the process clean regardless of who runs it
Whichever party ends up responsible for the actual collection, the underlying program fundamentals from SAMHSA’s workplace drug testing resources still apply: consistent application of the policy, documentation that can stand up if questioned later, and a collection process that doesn’t introduce its own weak points. An observed, straightforward collection method like a cotton swab drug test keeps the chain of custody simple regardless of which organization is administering it, which matters when a worker’s documentation may need to move between an agency’s files and a client company’s files without anything getting lost or disputed along the way.
Getting this sorted out before the first placement, rather than after a dispute, is what actually protects both the staffing agency and the business that relies on it.


