Objective Despite the efficacy of clinical treatments (eg, adjuvant hormonal therapy) for breast cancer survivors (BCS), nonadherence rates remain high, increasing the risk of recurrence and mortality. The current study tested a theoretical model of medical nonadherence that proposes depression to be the most proximal predictor of medical nonadherence among BCS. Methods Breast cancer survivors were recruited from radiation clinics in Missouri. Survey data were collected 12 months after the end of primary treatment. The sample size included 133 BCS. Results Findings show substantial support for the model, demonstrating that depression mediated the relation between physical symptoms, cognitive symptoms, social support, and adherence to medication. This finding was replicated with a measure of mood disturbance. Conclusions These findings support the prediction that medication nonadherence among BCS multiply determined process and have compelling implications for healthcare providers and interventions designed to increase medication adherence among BCS.