Objective Self-care is a key component in the management of chronic heart failure (CHF). Yet there are many barriers that interfere with a patient's ability to undertake self-care. The primary aim of the study was to test a conceptual model of determinants of CHF self-care. Specifically, we hypothesized that cognitive function and depressive symptoms would predict CHF self-care. Methods Fifty consecutive patients hospitalized with CHF were assessed for self-care (Self-Care of Heart Failure Index), cognitive function (Mini Mental State Exam), and depressive symptoms (Cardiac Depression Scale) during their index hospital admission. Other factors thought to influence self-care were tested in the model: age, gender, social isolation, self-care confidence, and comorbid illnesses. Multiple regression was used to test the model and to identify significant individual determinants of self-care maintenance and management. Results The model of 7 variables explained 39% (F [7, 42] 3.80; P = .003) of the variance in self-care maintenance and 38% (F [7, 42] 3.73; P = .003) of the variance in self-care management. Only 2 variables contributed significantly to the variance in self-care maintenance: age (P < .01) and moderate-to-severe comorbidity (P < .05). Four variables contributed significantly to the variance in self-care management: gender (P < .05), moderate-to-severe comorbidity (P < .05), depression (P < .05), and self-care confidence (P < .01). When cognitive function was removed from the models, the model explained less of the variance in self-care maintenance (35%) (F [6, 43] 3.91; P = .003) and management (34%) (F [6, 43] 3.71; P = .005). Conclusion Although cognitive function added to the model in predicting both self-care maintenance and management, it was not a significant predictor of CHF self-care compared with other modifiable and nonmodifiable factors. Depression explained only self-care management.