Stark Law is not a clinical-trial statute. It is a general physician self-referral rule that happens to reach into clinical research whenever a sponsor, CRO, or institution pays an investigator who is also a physician. Because that overlap is easy to miss, institutions sometimes address Stark Law only as a footnote inside broader clinical-trial billing or budget guidance — and then are surprised when a routine investigator agreement turns out to be a compensation arrangement the law actually regulates. This guide is scoped specifically to that overlap: what Stark Law prohibits, why it applies to investigator compensation at all, which exceptions institutions actually rely on to structure compliant agreements, and how Stark differs from — and interacts with — the federal Anti-Kickback Statute (AKS). For the separate question of how the False Claims Act attaches to clinical-trial billing itself (double-billing Medicare and a sponsor for the same routine cost, miscoded items), see False Claims Act Liability in Clinical Trial Billing. For negotiating the budget an investigator agreement sits inside, see Clinical Trial Budget Negotiation.
What Stark Law actually prohibits
The Physician Self-Referral Law — universally known as the Stark Law after its original congressional sponsor — is codified at 42 U.S.C. § 1395nn. In its core form, it prohibits a physician from referring Medicare (and, through a parallel Medicaid provision, Medicaid) patients for certain “designated health services” (DHS) to an entity with which the physician or an immediate family member has a financial relationship, unless that relationship fits squarely within a regulatory exception. Designated health services is a defined, enumerated list — it includes categories such as clinical laboratory services, radiology and certain imaging services, physical therapy, and inpatient/outpatient hospital services, among others set out in the statute and its implementing regulations at 42 CFR Part 411.
Three features of Stark Law make it unusually unforgiving compared to most fraud-and-abuse statutes:
- It is civil, not criminal. There is no fine or prison exposure under Stark itself; liability runs through denial of payment, refund obligations, and civil monetary penalties, and Stark violations are also a common predicate for False Claims Act liability when a tainted claim is submitted to Medicare.
- It is strict liability. No intent to violate the law, and no intent to induce referrals, is required. If a disqualifying financial relationship exists and a referral for DHS is made without the arrangement fitting an exception, a technical violation exists — full stop. This is the single most important difference from the Anti-Kickback Statute, covered below.
- It only reaches physicians’ own referrals, not referrals made by anyone else. A hospital paying a non-physician for referrals, or a physician’s referral of a non-Medicare/Medicaid patient, is outside Stark’s scope (though it may still implicate the Anti-Kickback Statute or state law).
Why Stark Law applies to clinical trial investigators at all
A sponsor or CRO paying a principal investigator to conduct a clinical trial is not, on its face, paying for a patient referral. But the financial-relationship trigger under Stark is broad: it is any compensation arrangement between a physician and an entity, and “entity” reaches the hospital, health system, or academic medical center that employs or credentials the investigator and that also furnishes DHS to Medicare beneficiaries — which is true of essentially every academic medical center and most hospital-based research sites. Once that financial relationship exists, the institution has to be able to show the arrangement fits a Stark exception, regardless of whether anyone involved thought of it as a “referral” arrangement in the colloquial sense.
The specific risk pattern compliance offices watch for is compensation structured — even inadvertently — in a way that varies with the volume or value of referrals the investigator generates for the institution’s other clinical services. A per-subject payment for actually performing trial-related visits, procedures, and data collection is not, by itself, a referral payment. A payment that increases because a physician enrolls their own existing patients, or that is set without reference to the actual work performed, looks much more like the kind of arrangement Stark exists to catch. The distinction is between paying an investigator for their labor and paying a physician because of who they refer — and the exceptions discussed next are built entirely around drawing that line.
The exceptions that actually apply to investigator agreements
There is no Stark exception written specifically for clinical trials or investigator compensation. In practice, institutions structure investigator agreements to fit one of two general compensation exceptions at 42 CFR § 411.357:
- Personal Service Arrangements, 42 CFR § 411.357(d). Requires that the arrangement be set out in a signed writing specifying the services covered; that it cover all services the physician furnishes to the entity (or clearly cross-reference other agreements that do); that its term run at least one year (or that it meet an alternative “deeming” standard for renewal); that aggregate compensation be set in advance, not exceed fair market value, and not be determined in any manner that takes into account the volume or value of referrals or other business generated between the parties; and that the arrangement be commercially reasonable even if no referrals were ever made.
- Fair Market Value Compensation, 42 CFR § 411.357(l). A more flexible exception that does not carry the same one-year minimum term, useful for shorter or single-study investigator agreements, but which still requires fair market value compensation set in advance and unrelated to referral volume or value, in a signed writing.
Both exceptions were reshaped by CMS’s 2020 Stark Law rulemaking (finalized December 2, 2020, effective January 19, 2021), which clarified that “set in advance” is a compliance-friendly deeming provision rather than a rigid pre-signature requirement and expanded what the fair-market-value exception could cover. That rulemaking did not create a clinical-trial-specific pathway — it made the existing general exceptions somewhat easier to satisfy correctly.
The recurring compliance failure point is fair market value itself. “FMV” for investigator compensation is not a single number; institutions typically benchmark against published clinical-trial-compensation survey data and document the basis for per-visit, per-procedure, or hourly rates so the file can demonstrate the rate reflects the actual scope of investigator work, not an estimate of what the trial is “worth” to the institution. Documentation of that benchmarking, alongside the signed agreement itself, is usually the first thing an auditor or an institution’s own compliance office will ask to see.
Stark Law vs. the Anti-Kickback Statute in clinical research
Stark Law and the federal Anti-Kickback Statute (AKS, 42 U.S.C. § 1320a-7b(b)) are frequently discussed together because they overlap heavily in clinical research fact patterns, but they are structurally different laws, and conflating them leads to compliance gaps:
- Criminal vs. civil. AKS is a criminal statute — knowingly and willfully offering, paying, soliciting, or receiving remuneration to induce referrals of anything reimbursable by a federal health care program is a felony. Stark is civil.
- Intent. AKS requires proof of knowing and willful intent to induce referrals. Stark requires none — an arrangement can violate Stark even when everyone involved acted in good faith and had no referral-inducing intent whatsoever.
- Who and what it reaches. Stark only restricts physicians’ self-referrals for the enumerated designated health services paid by Medicare/Medicaid. AKS is far broader: it applies to any individual or entity, covers any federal health care program (not just Medicare/Medicaid), and is not limited to physicians or to a defined list of services.
- Exceptions vs. safe harbors. Fitting a Stark exception is mandatory — an arrangement that doesn’t fit one is a violation. AKS “safe harbors” (42 CFR § 1001.952) are voluntary: failing to satisfy a safe harbor does not by itself mean the arrangement is illegal, only that it loses the automatic protection and must be evaluated case by case against the statute’s intent standard. The AKS counterpart to Stark’s personal-service-arrangements exception is the Personal Services and Management Contracts safe harbor at 42 CFR § 1001.952(d), which imposes largely parallel conditions: a signed writing, a term of at least one year, aggregate compensation set in advance at fair market value, and no variation based on referral volume or value.
Because the two laws share the same underlying fact pattern — a physician being paid by an entity that also benefits from that physician’s referrals — institutions generally structure a single investigator agreement to satisfy both the applicable Stark exception and the parallel AKS safe harbor at once, rather than treating them as separate compliance exercises. Neither law currently has a dedicated clinical-trial-specific exception or safe harbor; HHS’s Office of Inspector General has solicited public input on whether new safe harbors tailored to clinical trial participation are warranted, which signals this is an area regulators consider unsettled rather than fully resolved.
A related, narrower risk worth naming explicitly: paying a physician a fee for each patient referred into a study — a “finder’s fee” tied to enrollment numbers rather than to the investigator’s actual clinical-trial work — is the pattern most likely to be characterized as a kickback rather than legitimate investigator compensation, even where the underlying activity is a bona fide clinical trial. Compensation for the physician’s own time conducting trial-related visits, assessments, and documentation is a fundamentally different arrangement from compensation that scales with how many patients they route into the study.
How this fits into broader clinical-trial compliance
Stark Law and AKS analysis for investigator agreements does not happen in isolation. It typically sits alongside:
- Medicare coverage analysis under the qualifying-trial standard at NCD 310.1, which determines what costs Medicare will pay for in a clinical trial in the first place — see Medicare Clinical Trial Policy (NCD 310.1) and Medicare Coverage Analysis for Clinical Trials.
- False Claims Act exposure where a Stark or AKS violation taints a claim submitted to Medicare — see False Claims Act Liability in Clinical Trial Billing.
- Physician Payments Sunshine Act reporting, which is a separate transparency obligation (public disclosure via CMS Open Payments) rather than a self-referral or anti-kickback rule, but covers much of the same investigator-payment data — see Physician Payments Sunshine Act & CMS Open Payments.
- Institutional conflict-of-interest disclosure processes for investigators receiving industry compensation — see Conflict of Interest in Human Subjects Research.
- The compliance-program infrastructure (policies, training, auditing, reporting channels) that catches these issues before they reach a regulator — see The Seven Elements of an Effective Compliance Program, Applied to Research Compliance.
None of these substitute for Stark/AKS analysis specifically — an institution can have clean coverage analysis and full Sunshine Act reporting and still have a non-compliant investigator compensation arrangement if fair-market-value documentation or the “not tied to referrals” requirement was never addressed.
Frequently asked questions
Does Stark Law apply if the clinical trial has nothing to do with Medicare patients?
Stark’s referral prohibition is triggered by referrals of Medicare (or Medicaid) patients for designated health services. If an investigator’s institution furnishes DHS to Medicare beneficiaries at all — which is true of nearly every hospital-based or academic medical center research site, regardless of whether the trial itself enrolls Medicare patients — the investigator’s compensation arrangement with that institution is still a Stark-relevant financial relationship that needs to fit an exception. The safest practice is to treat every institutional investigator agreement as Stark-relevant rather than assuming a purely commercial or non-Medicare trial is exempt.
Is there a Stark Law exception written specifically for research payments?
No. Investigator compensation is analyzed under the general compensation exceptions at 42 CFR § 411.357, most commonly the Personal Service Arrangements exception (411.357(d)) or the Fair Market Value Compensation exception (411.357(l)), not a research-specific carve-out.
What is the single biggest documentation gap institutions have here?
Fair-market-value support for the specific rate paid — most commonly, an agreement whose per-visit or per-procedure rate isn’t tied back to a benchmarking source or documented rationale, which leaves the institution unable to demonstrate the “does not exceed fair market value” and “not determined by reference to referrals” elements if the arrangement is ever reviewed.
If an arrangement satisfies the Stark exception, is it automatically compliant with the Anti-Kickback Statute too?
Not automatically, though in practice the two are usually structured together. Satisfying a Stark exception addresses the civil self-referral prohibition; it does not by itself establish that the arrangement fits an AKS safe harbor, since the AKS analysis also asks about intent. Institutions generally design investigator agreements to meet the parallel safe harbor at 42 CFR § 1001.952(d) at the same time specifically to close this gap.
