The en banc Eleventh Circuit in Bolton v. Inland Fresh Seafood Corporation of America, Inc. overruled Mason v. Continental Group, Inc. “to the extent [Mason and its progeny] hold that ERISA itself requires exhaustion of statutory claims (including those alleging a breach of fiduciary duty).” The term “exhaustion” simply means that, before suing, a participant must first pursue the claim through the plan’s own internal claims and appeals process and give plan fiduciaries a chance to resolve it. The decision ends the Eleventh Circuit’s run as the only circuit to impose a mandatory exhaustion requirement on fiduciary breach claims, and it revives an ESOP class action that had been dismissed solely for failure to exhaust.
The early headlines say that exhaustion is no longer a defense to fiduciary breach claims in the Eleventh Circuit, normalizing the circuit with several other circuits. That overstates the decision. Bolton was careful to limit its holding to whether ERISA itself requires exhaustion of statutory claims, such as an alleged breach of fiduciary duty, and in a footnote expressly declined to address the Eleventh Circuit’s precedent requiring exhaustion where a plaintiff challenges the denial of benefits under the terms of a plan. Notably, the precedent Bolton cited for that line of cases does not speak in terms of benefit denials alone. Instead, it describes the requirement as applying to “ordinary breach-of-contract ERISA actions,” that is, claims that arise from the terms of the plan. Read that way, Bolton reinforces what appears to be the better reading of the circuit landscape: exhaustion should turn not on the label a plaintiff attaches to a claim, but on whether the claim, as a threshold matter, is based solely on statutory rights or whether it requires, in whole or in part, interpretation or application of plan terms. That distinction matters for plan sponsors and can be put to work through careful plan drafting.
What Bolton Decided
The plaintiffs, five former Inland Fresh employees suing on behalf of the company’s Employee Stock Ownership Plan (“ESOP”) and a proposed class of its 578 participants, alleged breaches of fiduciary duties including that plan fiduciaries caused the ESOP to overpay by tens of millions of dollars when it bought all of the company’s stock in 2016, and that the trustee failed to conduct adequate due diligence. The district court dismissed the complaint for failure to exhaust under Mason, and in 2025 a panel affirmed, with two judges concurring to urge the full court to overrule Mason. The court vacated the panel decision, received amicus briefing from the Department of Labor (“DOL”), heard en banc argument in June, and has now done exactly that.
The court’s reasoning is straightforward. ERISA’s text contains no exhaustion requirement for statutory claims, and Mason’s rule was, in the court’s words, “an atextual product of judicial creation based solely on policy reasons.” The court noted that the Third, Fourth, Fifth, Sixth, Ninth, Tenth, and D.C. Circuits do not require exhaustion of statutory claims, and that the Seventh Circuit, the only circuit to align somewhat with Mason, treats exhaustion as discretionary rather than mandatory. The court also declined to preserve Mason as a matter of judicial discretion, citing ERISA’s stated policy of “ready access to the Federal courts” and the concern that fiduciaries accused of malfeasance “may well have a conflict of interest in administratively reviewing a complaint against them.” The holding applies retroactively, including to the Bolton parties. Two other parts of the opinion deserve attention.
- First, the court did not decide whether a plan can contractually require exhaustion of statutory claims where ERISA is silent. It held only that this plan did not, because its claims procedure was permissive: a participant “need not file a claim” and “may submit a written claim,” and the summary plan description told participants they “may file suit in a federal court” if fiduciaries misuse plan money.
- Second, the court set aside the exhaustion language in the plan’s arbitration clause because no one had moved to compel arbitration, while noting its recent decision in Williams v. Shapiro declining to enforce an individual arbitration provision as a prospective waiver of ERISA rights, and the DOL’s statement at oral argument that contractual exhaustion requirements for statutory claims would face “an awful lot of judicial scrutiny.”
Bolton reversed the dismissal and remanded the case to the panel for any further proceedings.
The Line Bolton Preserves: Statutory Claims Versus Plan-Interpretation Claims
Cases like Zipf v. AT&T (3d Cir. 1986), Smith v. Sydnor (4th Cir. 1999), and Hitchcock v. Cumberland University 403(b) DC Plan (6th Cir. 2017) are often cited, and Bolton now joins them, for the proposition that fiduciary breach claims need not be exhausted. But those cases are better understood more narrowly than they are sometimes cited. They reject exhaustion for true statutory claims, meaning claims that seek to enforce ERISA-imposed duties of loyalty, prudence, monitoring, or prohibited-transaction rules and that do not depend on deciding what the plan means. Zipf drew the line between actions brought “not to enforce the terms of a plan, but to assert rights granted by the federal statute.” Hitchcock, which Bolton quotes, held that participants need not exhaust “when they assert statutory violations of ERISA,” while also making clear that the exception does not reach “plan-based claims artfully dressed in statutory clothing.”
In other words, these cases do not stand for the broader proposition that plaintiffs can avoid exhaustion simply by styling as a fiduciary breach what is really a plan administration or plan interpretation dispute. A good example is Harrow v. Prudential Insurance Co. of America (3d Cir. 2002), where the Third Circuit required exhaustion because the fiduciary-duty theory was bound up with whether the plan covered the disputed benefit, explaining that the claim “require[d] an interpretation and application of an ERISA-regulated plan rather than an interpretation and application of ERISA.”
Bolton fits comfortably within that framework. The court expressly confined its holding to “whether ERISA requires administrative exhaustion for statutory claims” and said it was not addressing precedent requiring exhaustion where a plaintiff challenges the denial of benefits under plan terms. The court also endorsed the Third Circuit’s observation that ERISA’s claims-procedure provision “refers only to procedures regarding claims for benefits.” And the secondary authorities the court cited in support of overruling Mason draw this same line. One treatise explains that fiduciary breach suits generally need not be exhausted “[b]ecause suits alleging breach of fiduciary duty involve statutory application rather than plan interpretation, [so] the special expertise of plan administrators with respect to plan language and prior applications would be of little or no assistance to a reviewing court.” Another states that exhaustion is not a prerequisite to remedying a statutory violation “if the violation is independent of the correct interpretation of the plan’s terms.” The flip side follows naturally: where the alleged violation is not independent of the plan’s terms, the rationale for skipping the plan’s process falls away.
The practical rule is straightforward. If a court cannot resolve the claim without first deciding what the plan says, the claim belongs in the plan’s administrative process first, especially where the plan grants fiduciaries discretion to interpret its terms and requires disputes over plan language to be exhausted. That rule has familiar limits. Courts continue to excuse exhaustion where it would be futile or where the claims procedure is inadequate, and under the DOL claims procedure regulation a claimant is generally deemed to have exhausted administrative remedies if the plan fails to establish or follow compliant claims procedures. Plan sponsors that want to be in a position to rely on exhaustion therefore need claims procedures that are not only mandatory on paper but also administered by plan fiduciaries in compliance with the regulation.
That approach is consistent with Supreme Court doctrine. Under Firestone Tire & Rubber Co. v. Bruch, courts review benefit determinations de novo unless the plan gives the administrator discretionary authority to determine eligibility or construe plan terms, in which case deferential review applies. That framework rests on the premise that the plan fiduciary should interpret the plan in the first instance. This makes sense for many reasons, including that interpretations should be applied consistently, not just in a particular case but across the plan, and a court simply will not, and should not, look over the shoulders of plan fiduciaries every day as they interpret plan provisions. To be sure, Firestone is a standard-of-review decision, not an exhaustion decision. Further, with respect to insured plans, some states prohibit discretionary clauses in insurance policies, meaning the deference piece of this argument may be weaker for those plans. But the premise that the fiduciary interprets the plan first holds regardless of the standard of review. Metropolitan Life Insurance Co. v. Glenn does not change that premise. Glenn treats a structural conflict of interest as a factor in judicial review; it does not eliminate the plan’s administrative process or prevent the fiduciary from making the initial interpretive decision.
Nor does Bolton’s conflict-of-interest concern undercut this approach. The court was concerned about fiduciaries “administratively reviewing a complaint against them” for malfeasance. Where the threshold question is what the plan permits or requires, the administrative process is not asking fiduciaries to adjudicate their own liability in the final instance. It is asking the fiduciary with interpretive authority to construe the plan, create a record, and apply the plan’s terms to adjudicate the plan’s liability, which is precisely the role Firestone contemplates.
This approach also need not be all or nothing. Just as Federal Rule of Civil Procedure 56 allows a party to move for summary judgment on “part of each claim or defense,” a court can separate a plan-interpretation issue embedded in a fiduciary claim from the statutory issues around it, for example by staying the statutory claims or the relevant portion of the case while the plan-interpretation issue is presented to the plan. That has implications beyond the pleadings. The plan-interpretation issue would be decided on the administrative record, with deference to the fiduciary where the plan grants discretion, while discovery and de novo analysis could proceed on the purely statutory questions. Put differently, a fiduciary would be entitled to deference on its interpretation of the plan, but not on its interpretation of ERISA.
How the Court on Remand Should Analyze Exhaustion
Bolton sent the case back to the panel, which may resolve any remaining issues itself or return the case to the Northern District of Georgia. Whichever court takes up the case, the opinion provides a clear road map, and it is one that tracks the statutory-versus-plan-interpretation framework described above. We would expect the analysis to proceed in four steps.
- Characterize the claim by its substance, not its label. Bolton does not hold that a fiduciary-breach caption ends the inquiry. Its holding is limited to statutory claims, and it expressly declines to address precedent requiring exhaustion of claims brought under the terms of a plan, which that precedent describes as “ordinary breach-of-contract ERISA actions.” The first question is therefore what the claim actually requires the court to decide, and in particular whether it requires interpretation or application of plan language.
- Ask whether a threshold plan-interpretation question exists. Here, the plaintiffs allege that fiduciaries caused the ESOP to overpay for company stock based on inflated projections and that the trustee failed to conduct adequate due diligence. Those are paradigm prudence, loyalty, and prohibited-transaction theories. As described in the opinion, resolving them requires measuring fiduciary conduct against ERISA’s standards, not construing the plan; the alleged violation is “independent of the correct interpretation of the plan’s terms.” Under Bolton’s holding, ERISA imposes no exhaustion requirement on those claims. If, however, any part of the case (or a case in the future) turned on a discrete plan term (for example, a dispute over how the plan itself directs a valuation or allocation), the court could require that issue to proceed through the plan’s process first, if required under the plan, consistent with the opinion’s first footnote and Hitchcock’s “artfully dressed” caveat, without requiring exhaustion of the statutory claims as a whole.
- Apply the plan’s actual exhaustion language. Bolton has already held that this plan’s claims procedure is permissive, reading it without the “Mason gloss” that previously led courts to treat similar language as merely reciting ERISA rights. That resolves any contractual exhaustion argument under the plan’s claims procedure for these claims, and the court on remand need not reach the open question of whether a plan may mandate exhaustion of statutory claims.
- Address arbitration only if it is properly raised. If a party moves to compel arbitration, the exhaustion language in the plan’s arbitration clause, Williams, and the DOL’s position on contractual waivers of statutory rights would come into play. Absent such a motion, Bolton says that language is not relevant.
Applied this way, the plaintiffs’ statutory claims seem likely to proceed on the merits without an exhaustion bar. That result is not in tension with the approach advocated here, instead it is exactly what that approach predicts for a claim that does not depend on plan interpretation. The more important point for plan sponsors is what the court should not do: treat Bolton as a categorical rule that any claim pleaded as a fiduciary breach escapes the plan’s administrative process, regardless of whether it turns on plan language. That point should be tested in appropriate future cases in which plaintiffs layer fiduciary breach claims on top of allegations that depend on how a plan’s terms were applied.
A Drafting Strategy Plan Sponsors Can Use Now
Bolton is also a drafting lesson. The court left open whether a plan can require exhaustion of statutory claims, but it showed how easily permissive language (“need not,” “may”) will be read as making the claims process voluntary. A strong exhaustion clause need not stop at denied benefit claims. At a minimum, it could use mandatory language and expressly cover claims alleging harm from interpretation of the plan, application of plan terms, and benefit calculations, as well as other claims. For example, the plan can require exhaustion for claims based on:
- interpretation of plan terms;
- application of eligibility, allocation, or distribution provisions;
- calculation or administration of benefits;
- action or inaction by the plan administrator, trustee, committee, or other plan fiduciary in administering the plan or applying its terms; and
- any alleged harm that depends on construing the plan.
The claims procedures could also make clear that exhaustion is required before filing any civil action to the fullest extent permitted by law, and the summary plan description could say the same thing. In Bolton, the summary plan description’s statement that participants “may file suit in a federal court” reinforced the court’s reading of the plan. Sponsors should also be sure the claims procedures satisfy the DOL claims procedure regulation and, if not already included, consider whether to include contractual limitations periods.
That drafting will not force every fiduciary breach claim through the claims procedure, nor should it; after Bolton and the DOL’s stated position, plan sponsors should expect courts to scrutinize any attempt to require exhaustion of purely statutory claims even more closely. Claims alleging imprudent investment selection, excessive recordkeeping fees, a conflicted transaction, or self-dealing may not turn on plan interpretation at all. But a well-drafted clause gives the plan sponsor a stronger, more precise argument where the plaintiff’s theory depends on what the plan permits, requires, or prohibits.
Thompson Hine Takeaways
The Eleventh Circuit is no longer an outlier on mandatory exhaustion of fiduciary breach claims. But the majority rule it has now joined is frequently oversimplified, and it does not stand for the proposition that claims rooted in the interpretation of plan provisions can bypass the plan’s administrative process. Plan sponsors and fiduciaries may wish to consider the following:
- Review plan documents and summary plan descriptions for permissive claims-procedure language (“may,” “need not”) and consider amendments using mandatory language that covers plan-interpretation and plan-application disputes, not only denied benefit claims.
- Confirm that the plan grants fiduciaries discretionary authority to interpret plan terms, to preserve the argument that first-instance interpretations receive deference under Firestone, and confirm whether any applicable state law limits any insurer from exercising that discretion.
- In litigation, focus exhaustion arguments on the specific plan-interpretation issues embedded in a fiduciary claim, rather than on the fiduciary-breach label, and consider whether there may be strategic advantages in seeking exhaustion of those issues even where statutory issues proceed.
- If the plan includes an arbitration clause with exhaustion language, revisit it in light of Williams v. Shapiro and the DOL’s position on contractual limits on statutory claims.
Please contact the authors or your Thompson Hine attorney with any questions.