Differential effects on fitness are hypothesized to contribute to the opposing health effects of leisure-time physical activity (LTPA) and occupational physical activity (OPA). As such, this study examined cross-sectional and longitudinal associations of fitness with LTPA and OPA. This study examined fitness associations with LTPA and OPA across 13 years in the Coronary Artery Risk Development in Young Adults (CARDIA) Study (years 7 [baseline], 10, 15, and 20 [follow-up] exams). Fitness was measured at baseline and follow-up via symptom-limited maximal graded exercise test (GXT) duration (seconds) while LTPA and OPA were self-reported during each exam. Baseline and follow-up cross-sectional associations of LTPA (low, medium, high) and OPA (0, 1-6, and ≥ 6 months with OPA) with fitness were examined using linear regression. Longitudinal linear regression examined associations between 13-year LTPA (low, medium, or high) and OPA (no, decreasing, or increasing) trajectories with fitness at follow-up, adjusted for baseline values. All models adjusted for center, sex, race, age, education, smoking history, alcohol intake, resting blood pressure, diabetes status, and body mass index. Stratified analyses examined associations by sex (female/male), race (Black/White), and LTPA groups. Compared to low, medium and high LTPA were positively associated with fitness in all analyses (p < 0.001). Reporting 1-6 or ≥ 6 months with OPA was negatively associated with fitness in cross-sectional follow-up models (β = -15.6 and -15.4 respectively, p ≤ 0.01). Longitudinally, those with increasing OPA had lower follow-up fitness compared to no OPA (β = -16.41, p < 0.01). Negative associations of OPA with fitness were not meaningfully different across sex and race groups. Significant LTPA by OPA interactions were observed (p < 001). Physical activity research and public health promotion should consider domain-specific associations on cardiovascular health.