Major depression and coronary heart disease are two strongly linked, major causes of death and disability. After an acute coronary event, many patients are referred to cardiac rehabilitation (CR), a medically supervised exercise intervention and lifestyle training program. Depression may partially account for poor CR adherence and resulting cardiovascular problems in patients with a history of heart disease; however, underlying mechanisms through which depression impacts cardiac functioning are not well understood. The current project tests a theoretical model in which CR adherence (i.e., number of CR sessions attended) mediates the relation between baseline depression and cardiorespiratory fitness after CR. A community sample of 858 older adults initiating CR after hospitalization for a coronary event completed a symptom-limited exercise stress test before and after the 12-week program. Cardiorespiratory fitness was measured via VO2max, peak MET, and total duration of the stress test. Depression was measured at baseline using the Patient Health Questionnaire Depression Scale. CR adherence did not mediate the relation between baseline depression scores and fitness outcomes. Path analyses revealed that higher baseline depression severity predicted lower likelihood of CR completion (i.e., completion of all 36 sessions, or fewer if limited by insurance or terminated early for good prognosis) in the full sample. Higher levels of depression predicted poorer CR adherence in a subsample of 74 patients with moderate to severe depression. These findings lend support to depression as a predictor of treatment nonadherence in CR. Screening for depression in the context of coronary heart disease and implementing evidence-based depression interventions in secondary prevention settings can help alleviate a massive public health burden.