The current male-costs hypothesis to explain variation in the timing of pairing in waterfowl acknowledges that relative parental investment and intra-sexual competition among males place females in control of pairing in most species. However, because females may benefit from early pair formation, it assumes that they should be willing to pair whenever males are energetically capable, and thus predicts that timing of pairing depends on decisions made by males. A compilation and review of available data on pairing behavior in waterfowl found little support for this hypothesis and considerable data that were inconsistent with its predictions. The mutual-choice hypothesis extends the male-costs hypothesis by incorporating cost-benefit trade-offs to females as well as males, and by giving more consideration to the various components of the pairing process, including time and energy invested into mate choice, the quality of mate chosen, and the state of being paired, including possible benefits of time spent gaining familiarity with a partner before breeding. This hypothesis differs from the male-costs hypothesis in two essential points: 1) that decisions by females rather than males primarily determine the timing of pairing, and 2) that it is necessary to incorporate the process of choosing a partner as well as the state of being paired. The new hypothesis was supported by available data and by results of preliminary, comparative analyses, and provides a theoretical structure to describe phylogenetic trends in pairing behavior. However, critical testing of many predictions is currently hampered by a lack of longitudinal data on age- and sex-specific pairing chronologies using individual-based measures of pairing chronology.