COPD patients have reduced levels of physical activity (PA) and a high prevalence of anxiety and depression, but the relationship between these variables is poorly understood. Main objective: To assess the longitudinal effects of anxiety and depression on PA in COPD patients. Methods: We recruited 236 COPD patients from 5 European countries and followed them for 1 year (EU IMI PROactive project). Anxiety and depression symptoms (using the Hospital Anxiety Depression Scale), PA measures (with DynaPort ® accelerometer), socio-demographic characteristics, lung function tests, bioimpedance and exercise capacity (6MWD) were evaluated at baseline, and at 6 and 12 months. We built multivariate linear regression models with repeated measures to assess the association of anxiety/depression (at baseline and 6 months) with PA (at 6 and 12 m). Results: Patients (age 67(8)y, FEV1 57(21)%) showed a prevalence of depression and anxiety (HADS depression and anxiety scores≥11 points) of 6% and 12%, respectively. Patients with depression performed fewer steps/day (1894 vs 4143, p<0.01) and spent less time in locomotion (26 vs 56 min/day, p<0.01) than patients without depression. In multivariate models adjusted by confounders (age, FEV1%, 6MW distance, smoking) and baseline PA, per each extra point in HADS depression score, patients performed -101 (IC95% -189 to -13) steps/day and -1.1 (-2.1 to -0.1) min/day of time in locomotion. Anxiety symptoms were not associated with PA. Conclusions: In patients with COPD, symptoms of depression are associated with less PA six months later. No longitudinal effect of anxiety on PA was found.