An Institutional Review Board (IRB) is not a single reviewer or a compliance office — it is a committee, and federal regulation dictates who has to sit on it, in what numbers, and with what mix of backgrounds. Those composition rules exist for a specific reason: an IRB is supposed to weigh a study’s scientific merit and its risk to participants at the same time, and no single discipline is positioned to judge both. This guide explains exactly what 45 CFR 46.107 (the Common Rule) and 21 CFR 56.107 (FDA’s parallel provision for FDA-regulated research) require of IRB membership, what each required role actually does, and how composition connects to the board’s ability to convene and vote.
What Counts as an IRB Member
An IRB member is an individual formally appointed to a specific, institutionally registered board and listed on that board’s official roster — the same roster the institution submits to the Office for Human Research Protections (OHRP) as part of its Federalwide Assurance (FWA) registration. Membership is not informal: a researcher who occasionally advises a board, or a department chair who signs off on protocols administratively, is not an IRB member unless they hold a seat on that roster and participate in review and voting under the board’s own written procedures.
The Regulatory Basis: 45 CFR 46.107 and 21 CFR 56.107
IRB composition requirements come from two closely parallel federal provisions. 45 CFR 46.107, part of the Common Rule (45 CFR 46 Subpart A), governs IRBs reviewing federally funded human subjects research. 21 CFR 56.107 is FDA’s near-identical companion provision for IRBs reviewing research regulated under the Food, Drug, and Cosmetic Act (clinical trials of drugs, biologics, and devices — see 21 CFR Part 56). Most academic medical centers and universities conducting both federally funded and FDA-regulated research operate under both sets of requirements simultaneously, and in practice satisfy them with the same board and the same roster.
Minimum Size and Diversity Requirements
Under 45 CFR 46.107(a), every IRB must have at least five members, with varying backgrounds selected to promote complete and adequate review of the research the institution commonly conducts. The regulation is specific about what “varying backgrounds” means in practice:
- Professional competence. The board as a whole must be qualified through the experience and expertise of its members to evaluate the science, the ethics, and the institutional/legal context of the protocols it reviews.
- Diversity of members, including race, gender, and cultural background, and sensitivity to issues such as community attitudes — intended to promote respect for the board’s judgment among the populations affected by its decisions.
- No single-profession or single-gender board. Both 45 CFR 46.107(a) and 21 CFR 56.107(a) require that every reasonable, nondiscriminatory effort be made to ensure a board does not consist entirely of men, entirely of women, or entirely of members of one profession.
- Legal and institutional literacy. The board must include people knowledgeable enough about applicable law, institutional commitments, and standards of professional conduct to judge whether a proposed study is acceptable on those grounds, not only on scientific grounds.
The Required Member Categories
Beyond overall diversity, three specific categories of membership are mandatory. An institution can staff a board well past five members, but it cannot legally convene one without covering these three roles.
At least one scientist member
45 CFR 46.107(b) requires at least one member whose primary concerns are in scientific areas — someone equipped to assess study design, methodology, and the scientific justification for the risks a protocol poses. In practice, most boards seat several scientist members spanning the institution’s major research areas (biomedical, behavioral, social science) so that reviewers with relevant subject-matter expertise are typically available.
At least one non-scientist member
The same subsection requires at least one member whose primary concerns are in nonscientific areas — for example, an ethicist, clergy member, lawyer, or someone from a humanities or arts background. The non-scientist member’s function is to keep the board’s deliberation from becoming purely a technical/methodological conversation and to represent perspectives outside the research disciplines under review.
At least one non-affiliated member
45 CFR 46.107(c) requires at least one member who is not otherwise affiliated with the institution and who is not an immediate family member of someone who is affiliated with it. This member represents the community and general public interest rather than the institution’s own research enterprise, and is meant to counterbalance any institutional bias in the board’s decisions.
Can one person satisfy more than one category?
Yes. The regulation defines categories of required expertise and perspective, not a fixed head count of distinct individuals — a single member can, for example, simultaneously be the board’s non-scientist member and its non-affiliated member, provided they genuinely meet both definitions. Many institutions deliberately combine these categories in one seat because qualified non-affiliated community members with a nonscientific background are easier to recruit and retain than filling two separate seats. What is not permissible is skipping a category outright because a similar-sounding role is filled — a non-affiliated scientist does not satisfy the non-scientist requirement, for instance.
Expertise for Vulnerable Subject Populations
45 CFR 46.107(a) also directs institutions to consider the inclusion of members experienced in working with vulnerable categories of subjects — commonly cited examples include children, prisoners, and individuals with impaired decision-making capacity or who are economically or educationally disadvantaged — when the IRB regularly reviews research involving those populations. This is a standing consideration built into board composition, separate from (and in addition to) the additional procedural safeguards those populations trigger under other Common Rule subparts, such as 45 CFR 46 Subpart C for research involving prisoners.
Alternate Members
Many institutions formally designate alternate members who can stand in for a primary member who is absent or recused from a given review. This is a common, OHRP-consistent institutional practice rather than a distinct federal head-count requirement in its own right: where an institution uses alternates, the expectation is that an alternate for a specific required role (the non-scientist seat, or the non-affiliated seat) must independently meet that same role’s qualifications — an alternate cannot fill in for the non-affiliated member unless that alternate is also non-affiliated. Institutions document alternate-member assignments in the same IRB roster and standard operating procedures used to satisfy the core composition requirements.
Ad Hoc Consultants and Non-Voting Experts
Under 45 CFR 46.107(e) / 21 CFR 56.107(f), an IRB may invite individuals with competence in special areas to assist in reviewing issues that fall outside the expertise already represented on the board — a pediatric specialist for a study involving children, a statistician for a complex trial design, or a community representative with lived experience relevant to the population being studied. These consultants can contribute to discussion but may not vote with the IRB; only seated members exercise the board’s voting authority.
Conflict of Interest: When a Member Must Step Back
45 CFR 46.107(d) prohibits any IRB member from participating in the initial or continuing review of a project in which that member has a conflicting interest — for example, being an investigator on the study, having a direct financial stake in its outcome, or another relationship that could bias review — except to provide information the board specifically requests. This is a review-level recusal mechanism distinct from an institution’s broader financial conflict-of-interest management process for investigators; see Conflict of Interest in Human Subjects Research for how the two intersect. A recused member does not count toward quorum for that specific agenda item.
How Composition Connects to Quorum and Voting
Composition requirements are not just a staffing checklist — they determine whether a board can legally act. Under 45 CFR 46.108, except when expedited review applies, an IRB may only review research at a convened meeting with a majority of its members present, and that quorum must include at least one member whose primary concerns are in nonscientific areas. A study can only be approved by a majority vote of members present at that meeting, and can only be disapproved through this convened-meeting process — not through expedited review. That means a board that is technically fully staffed on paper but cannot physically convene a quorum including its non-scientist member cannot legally approve or disapprove full-board studies that day. See IRB/REC Approval Process for how review type and voting fit together, and the IRB Chair entry for who presides over that meeting and confirms quorum.
Documenting the Roster: OHRP Registration and the FWA
Institutions register each of their IRBs with OHRP, and that registration includes a current roster naming the Chair, Vice Chair (where applicable), and all voting members, alternates, and consultants, along with each member’s role, department/affiliation, and area of expertise. Registration is tied to the institution’s Federalwide Assurance (FWA) — the written, time-limited institutional commitment to comply with the Common Rule. Keeping the roster current (updating it when members join, leave, or change roles) is an ongoing administrative responsibility of the IRB Coordinator or equivalent HRPP staff, not a one-time filing.
IRB Member Roles at a Glance
Illustrative composite example, not drawn from any specific institution’s actual roster. A typical seven-member board that satisfies the required categories, with some members covering more than one role, might look like this:
- Two biomedical scientist members (the required scientist role, plus depth for the institution’s dominant research area)
- One social/behavioral scientist member
- One non-scientist member who is also non-affiliated with the institution (satisfying both the (b) and (c) requirements in a single seat)
- One community member with experience relevant to a vulnerable population the institution regularly studies
- One institutional research-compliance staff member serving as a scientist-category member with regulatory expertise
- One alternate member qualified to stand in for the non-scientist/non-affiliated seat when that member is recused or absent
The specific mix varies enormously by institution size and research portfolio; the regulatory floor is the five-member minimum plus the scientist, non-scientist, and non-affiliated categories described above — everything past that floor is an institutional design choice.
Frequently Asked Questions
What is the minimum number of members an IRB must have?
At least five, under both 45 CFR 46.107(a) and 21 CFR 56.107(a). There is no federal maximum; larger institutions with high protocol volume commonly run boards well above the five-member floor, or operate multiple IRBs, to manage caseload.
Does every IRB meeting need a non-scientist member present?
Yes for a convened, full-board review. 45 CFR 46.108 requires quorum at a convened meeting to include at least one member whose primary concerns are in nonscientific areas; without that member present, the board cannot approve or disapprove research requiring full-board review at that meeting.
Can the non-affiliated member also be the non-scientist member?
Yes, provided the individual genuinely meets both definitions. The regulation requires the categories to be covered, not necessarily by separate people, and many institutions combine these two roles in a single community-member seat.
Who decides how many members an IRB has beyond the five-member minimum?
The institution, through its own written IRB procedures required under 45 CFR 46.103(b) and 21 CFR 56.108. Federal regulation sets the floor and the required categories; board size, number of alternates, and additional scientist/non-scientist seats above that floor are institutional decisions based on protocol volume and research portfolio.
Do consultants who advise the IRB on a study count toward quorum?
No. Individuals invited under 45 CFR 46.107(e) to assist with specialized review issues may contribute to discussion but do not vote and are not counted as part of the quorum — only seated members and, where used, properly qualified alternates count.
