Proposed causal links between extra-pair copulation (EPC) and colony formation in socially monogamous birds hinge on the question of which sex controls fertilizations. We examined in colonial purple martins Progne subis (1) whether EPCs were forced or accepted by females, and (2) the degree to which apparently receptive females were able to obtain EPCs against their mates' paternity defenses, Paternity analyses of multilocus DNA fingerprinting confirmed previous findings of a marked relationship between age class and extra-pair fertilizations (EPFs), with young males losing paternity of 43% (n = 53) of their putative offspring compared to 4% (n = 85) by old males, All assignable extra-pair offspring were sired by old males, with one male obtaining most EPFs each year, Contrary to the hypothesis that EPCs are forced, EPF frequency within age class did not increase with seasonal increases in the number of males per fertile female. Whereas the male control hypothesis predicted that the male age class that mate-guarded more would be cuckolded less, the reverse was true: young males guarded significantly more intensely. The male age class difference in cuckoldry could not be explained by the possibility that young and inexperienced females (which are usually paired to young males) were more vulnerable to forced copulation because EPFs were unrelated to female age, These findings suggest that females (1) pair with old males and avoid EPCs, or (2) pursue a mixed mating strategy of pairing with young males and accepting EPCs from old males. The receptivity to EPCs by females paired to young males put them in conflict with their mates. Two factors determined the paternity achieved by young males: (1) the relative size of the male to the female, with young males achieving much higher paternity when they were larger than their mates, and (2) the intensity of mate-guarding, Both variables together explained 77% of the variance in paternity and are each aspects of male-female conflict. Given female receptivity to EPCs, mate-guarding can be viewed as male interference with female mating strategies. We conclude that EPCs are rarely or never forced, but the opportunity for females paired to young males to obtain EPCs is relative to the ability of their mates to prevent them from encountering other males, Evidence of mixed mating strategies by females, combined with other features of the martin mating system, is consistent with the female-driven ''hidden lek hypothesis'' of colony formation which predicts that males are drawn to colonies when females seek extrapair copulations.