To explore whether selection for testosterone-mediated traits in males might be constrained by costs of higher testosterone to females, we examined the effects of experimental elevation of plasma testosterone on physiological, reproductive, and behavioral parameters in a female songbird, the dark-eyed junco (Junco hyemalis). We used subcutaneous implants to elevate testosterone (T) in captive and free-living female juncos. In captive birds, we measured the effects of high T on body mass, feather molt, and brood patch formation. In the field, we monitored its effects on the timing of egg laying, clutch size, egg size, egg steroid levels, incubation, and nest-defense behavior. Females implanted with testosterone (T-females) had significantly higher circulating levels of testosterone than did control females (C-females). Captive T-females had lower body mass, were less likely to develop brood patches, and delayed feather molt relative to C-females. Among free-living females, the interval between nest completion and appearance of the first egg was longer for T-females than for C-females and egg yolk concentrations of testosterone were higher, but there were no significant differences in estradiol levels, clutch size, or egg size. Incubation and nest defense behavior were also similar between T- and C-females. Our results suggest that selection on males for higher testosterone might initially lead to a correlated response in females producing changes in body mass and feather molt, both of which could be detrimental. Other possible female responses would be delayed onset of reproduction, which might reduce reproductive success, and higher yolk testosterone, which might have either positive or negative effects on offspring development. We found no reason to expect reduced parental behavior by females as a negative fitness consequence of selection for higher testosterone in males.