The aim of this study is to investigate the effects of exercise as an add-on therapy with antidepressant medication and cognitive behavioral group therapy (CBGT) on treatment outcomes in low-active major depressive disorder (MDD) patients. We also explored whether exercise reduces the residual symptoms of depression, notably cognitive impairment and poor sleep quality, and aimed to identify putative biochemical markers related to treatment response. Sixteen low-active MDD patients were recruited from a mental health day treatment program at a local hospital. Eight medicated patients performed an 8-week exercise intervention in addition to CBGT, and eight medicated patients attended the CBGT only. Twenty-two low-active, healthy participants with no history of mental health illness were also recruited to provide normal healthy values for comparison. Results showed that exercise resulted in greater reduction in depression symptoms ( = 0.007, = 2.06), with 75% of the patients showing either a therapeutic response or a complete remission of symptoms vs. 25% of those who did not exercise. In addition, exercise was associated with greater improvements in sleep quality ( = 0.046, = 1.28) and cognitive function ( = 0.046, = 1.08). The exercise group also had a significant increase in plasma brain-derived neurotrophic factor (BDNF), = 0.003, = 6.46, that was associated with improvements in depression scores ( = 0.002, = 0.50) and sleep quality ( = 0.011, = 0.38). We provide evidence that exercise as an add-on to conventional antidepressant therapies improved the efficacy of standard treatment interventions. Our results suggest that plasma BDNF levels and sleep quality appear to be good indicators of treatment response and potential biomarkers associated with the clinical recovery of MDD.