Divorce is widespread among species of birds and may either be an adaptive strategy to secure a better mate or territory or be a non-adaptive result of a failure to maintain the pairbond. I examined the causes and consequences for divorce in the Norther Flicker (Colaptes auratus), a migratory woodpecker with a high annual mortality rate. In a long-term population study of 1793 breeding pairs over 17 years, the within-season divorce rate was 4.6% and the between-season divorce rate was 15.5%. Retained pairs within a season initiated their renest 5 days faster than divorced birds which had no greater fledgling production, suggesting that within-season divorce was making the best of a bad job with severe time constraints. Poor performance in the year prior to divorce was not strongly associated with divorce and analysis of multiple breeding stages revealed that divorcing individuals in the subsequent year had later laying dates, smaller clutches and fewer fledglings than retained pairs but no better performance than widowed individuals. Analyzing the data separately by sex showed that neither males nor females benefitted from divorce. Thus, there is a reproductive cost linked to finding a new partner per se, but no reproductive advantage associated with divorce. New mates after divorce were usually not older (not higher quality) than previous mates, so intrasexual competition was probably not driving partnership splits. The most plausible explanation seems to be a "bet-hedging" hypothesis in which birds re-pair rapidly in spring if their previous mate does not quickly arrive during spring migration. Divorce in Northern Flickers does not appear to be adaptive and future studies on arrival and interactions of individuals in spring will elucidate proximate constraints on relocating the previous partner.