Purpose. RCTs have demonstrated the benefits of exercise in heart failure (HF) including improved functional status and quality of life, with fewer hospitalizations. Little is known about how changes in physical activity (PA) over time may influence physical and psychological symptoms. This study examined the extent to which changes in PA over time were associated with changes in physical and psychological symptoms in HF. Methods. A secondary analysis was conducted using an RCT of exercise in HF patients (N=86) who received an initial ICD. Data were collected at baseline and 2 months. Physical activity was measured over 5 consecutive days using the StepWatch PA Monitor (Orthocare Innovations TM ) for total PA (mean steps/day) and intensity (mean steps/min for most intense 30 minutes/day). Other measures included: physical symptoms (Patient Concerns Assessment (PCA); fatigue and pain (SF-36 Vitality & Bodily Pain subscales, respectively), and symptoms of depression (Patient Health Questionnaire-9, PHQ9) and anxiety (State-Trait Anxiety Index, STAI). Multivariate regression was used to examine for associations between increased and decreased PA and changes in physical and psychological symptoms, adjusting for age, co-morbidities, body mass index, ejection fraction (EF), and study group. Results. Patients were male (82%), 55.3±12.1 years of age, white (79%), with multiple co-morbidities (Charlson Co-morbidity: 2.1±1.3), BMI of 30.0±5.1, with reduced EF (31.0±10.1%). An increase in PA intensity from baseline to 2 months was associated with significant improvement in psychological symptoms (PHQ9:β= -0.13, p=0.030; STAI:β= -0.26,p=0.026) and fatigue (β= 0.55,p=0.023). An increase in total PA (steps/day) was associated with improvement in anxiety only (β= -1.23, p=0.045). A decrease in PA (total and intensity) was not associated with changes in any symptoms. Conclusion. For patients with HF and an initial ICD, increased intensity of PA over 2 months was associated with improved psychological symptoms and reduced fatigue. Decreases in PA (total and intensity) were not associated with symptoms. These findings warrant further investigation as interventions focused on PA intensity rather than quantity may be more effective in reducing symptoms in HF.