BACKGROUND: Patients knowledge about heart failure (HF) contributes to successful HF self-care, but less is known about shared patient-caregiver knowledge. OBJECTIVES: The purpose of this analysis was to: 1) identify configurations of shared HF knowledge in patient-caregiver dyads; 2) characterize dyads within each configuration by comparing sociodemographic factors, HF characteristics, and psychosocial factors; and 3) quantify the relationship between configurations and patient self-care adherence to managing dietary sodium and HF medications. METHODS: This was a secondary analysis of cross-sectional data (N = 114 dyads, 53% spousal). Patient and caregiver HF knowledge was measured with the Atlanta Heart Failure Knowledge Test. Patient dietary sodium intake was measured by 3-day food record and 24 h urine sodium. Medication adherence was measured by Medication Events Monitoring System caps. Patient HF-related quality of life was measured by the Minnesota Heart Failure Questionnaire; caregiver health-related quality of life was measured by the Short Form-12 Physical Component Summary. Patient and caregiver depression were measured with the Beck Depression Inventory-II. Patient and caregiver perceptions of caregiver-provided autonomy support to succeed in heart failure self-care were measured by the Family Care Climate Questionnaire. Multilevel and latent class modeling were used to identify dyadic knowledge configurations. T-tests and chi-square tests were used to characterize differences in sociodemographic, clinical, and psychosocial characteristics by configuration. Logistic/linear regression were used to quantify relationships between configurations and patient dietary sodium and medication adherence. RESULTS: Two dyadic knowledge configurations were identified: Knowledgeable Together (higher dyad knowledge, less incongruence; N = 85, 75%) and Knowledge Gap (lower dyad knowledge, greater incongruence; N = 29, 25%). Dyads were more likely to be in the Knowledgeable Together group if they were White and more highly educated, if the patient had a higher ejection fraction, fewer depressive symptoms, and better autonomy support, and if the caregiver had better quality of life. In unadjusted comparisons, patients in the Knowledge Gap group were less likely to adhere to HF medication and diet. In adjusted models, significance was retained for dietary sodium only. CONCLUSIONS: Dyads with higher shared HF knowledge are likely more successful with select self-care adherence behaviors.