Abstract: Previous studies have shown that experimentally elevated testosterone (T) causes male dark-eyed juncos (Junco hyemalis) to be more attractive to females, occupy larger home ranges, and achieve more extra-pair fertilizations (EPFs).  Although elevated T decreases male parental care, females attempt to compensate for this loss by increasing their own maternal effort.  Based on attractiveness and EPF frequency, elevated T ought to be favored by sexual selection unless there is some counteracting force that constrains its evolution.  Occasionally, a trait that is advantageous for a male may be disadvantageous when it appears as a correlated trait in females.  In this study, we examined the effects of experimentally elevated T on mate choice in female juncos.  We used subcutaneous implants to increase T levels in both male and female juncos and performed trials in which a female chose between a control (C-) male and a testosterone-implanted (T-) male.  We compared the preference of control (C-) females to that of testosterone implanted (T-) females using a paired design.  We hypothesized that if T-females preferred C-males,  this altered preference might act to constrain the evolution of high T in males.  Unlike previous studies, we found that that C-females preferred C-males.  We also found that treatment with T made females indifferent in their mate choice.  Half of the T-females preferred T-males, while the other half preferred C-males.  We consider why our results differ from past experiments and conclude that the preference of C-females for C-males may have been affected by their having been housed with T-females.  If T-females acted more aggressively than C-females, the stressful conditions may have compelled C-females to make the “safer” choice when choosing a mate.  Also, the preference for C- or T-males may be contingent on male age.  We interpret the lack of choice made by T-females as lack of ability or willingness to choose a mate.  This property of T-females may act as an constraint, since females with high T in nature may be unable to choose a mate wisely and thus have a competitive disadvantage in attaining high quality males as social mates.  To our knowledge,  this is the first study of a wild a passerine species to show that a hormonal correlation between the sexes could constrain sexually selected traits in males.