BACKGROUND Exercise improves cardiorespiratory fitness (CRF) and reduces depressive symptoms in people with depression. It is unclear if changes in CRF are a predictor of the antidepressant effect of exercise in people with depression. OBJECTIVE To investigate whether an increase in CRF is a predictor of depression severity reduction after 12 weeks of exercise (trial registration: DRKS study ID, DRKS00008745). METHODS The present study includes participants who took part in vigorous (n = 33), moderate (n = 38) and light (n = 39) intensity exercise and had CRF information (as predicted maximal oxygen uptake, V̇O2max) collected before and after the intervention. Depression severity was measured with the Montgomery-Åsberg Depression Rating Scale (MADRS). V̇O2max (L/min) was assessed with the Åstrand-Rhyming submaximal cycle ergometry test. The main analysis was conducted pooling all exercise intensity groups together. RESULTS All exercise intensities improved V̇O2max in people with depression. Regardless of frequency and intensity of exercise, an increase in post-treatment V̇O2max was significantly associated with reduced depression severity at follow-up (B = -3.52, 95% CI -6.08 to -0.96); adjusting for intensity of exercise, age and body mass index made the association stronger (B = -3.89, 95% CI -6.53 to -1.26). Similarly, increased V̇O2max was associated with higher odds (odds ratio = 3.73, 95% CI 1.22-11.43) of exercise treatment response (≥50% reduction in MADRS score) at follow-up. CONCLUSIONS Our data suggest that improvements in V̇O2max predict a greater reduction in depression severity among individuals who were clinically depressed. This finding indicates that improvements in V̇O2max may be a marker for the underpinning biological pathways for the antidepressant effect of exercise. BACKGROUND None.