The "challenge hypothesis" posits that when established social orders are challenged, plasma testosterone (T) in socially monogamous breeding male birds will temporarily increase to facilitate aggressive responses. However, not all birds conform to predictions. To expand upon past findings, we examined effects of direct territorial challenge on T levels in the blue tit (Cyanistes caeruleus). We found that simulated territorial intrusions caused a decline in plasma T during both territory establishment and laying/incubation. Conversely, corticosterone (CORT) levels dramatically increased. We also examined challenged blue tit males for levels of corticosterone-binding globulin (CBG), a carrier molecule that displays affinity for both CORT and T in birds. Although the CBG showed increased occupation by CORT during challenge, effects were not accompanied by a significant increase in the unbound T fraction. Thus, competitive hormone interactions on the CBG do not seem sufficient to explain changes in circulating T levels. To place our results within the context of past findings, we compared all socially monogamous birds tested to date for plasma levels of T during situational territorial intrusion experiments. We found that birds raising only one brood per season (e.g., the blue tit) consistently show no increase in plasma T but instead show elevations in circulating CORT. Thus, we suggest that single-broodedness plays an important role in determining patterns of hormone change and should be considered in future discussions of hormone-behavior interactions.