Each breeding season, seasonally monogamous birds can divorce or reunite with their previous year's mate, assuming both partners survive and return. We tested a suite of variables related to mate choice and site choice to determine which of 4 prominent mate fidelity hypotheses (better [mate] option, habitat mediated, musical chairs, and bet-hedging) best explained the interyear reunion rate and breeding dispersal of a seasonally monogamous shorebird, Piping Plover (Charadrius melodus), nesting on dynamic sandbars on the Missouri River, USA, 2005–2012. Of 252 pairs in which both members returned to the breeding grounds the following year, only 20% reunited. Pairs with an early-arriving male had the highest mate fidelity. Reunited pairs returned to previous nest sites (median breeding dispersal = 39 m), and divorced and widowed birds tended to move farther (median = 229 m and 193 m, respectively). Overall, site fidelity was higher in males than females. Previous reproductive success of a pair did not predict reunion, but all successful birds, with the exception of divorced females, exhibited high site fidelity, suggesting selection for site based on prior breeding success. Among divorced birds, females had higher-quality mates and higher nest success compared to their former partners, and they nested in areas of similar quality between years, whereas males settled in lower-quality areas following divorce. The benefits that females gained from divorce suggested that females initiated divorce to improve reproductive success, which supports the better option hypothesis. Although females seemed to initiate most divorces, males may have divorced as a safeguard against remaining unmated when there was uncertainty about the survival and return of a former mate, as proposed by the bet-hedging hypothesis.