Mediating mechanisms of a 12-week group-based exercise intervention on cancer survivors’ quality of life (QoL) were examined to inform future exercise intervention development. Methods: Two hundred nine cancer survivors ≥3 months post treatment (57% breast cancer) aged 49.5 (±10.4) years were assigned to physical exercise (n = 147) or wait-list control (n = 62). QoL, fatigue, emotional distress, physical activity, general self-efficacy and mastery were assessed at baseline and post-intervention using questionnaires. Path analysis was conducted using Mplus to explore whether improved physical activity, general self-efficacy and mastery mediated the effects of exercise on fatigue and distress and consequently QoL. Results: The intervention was associated with increased physical activity (β = 0.46, 95% confidence interval (CI) = 0.14;0.59), general self-efficacy (β = 2.41, 95%CI = 0.35;4.73), and mastery (β = 1.75, 95% CI = 0.36;2.78). Further, the intervention had both a direct effect on fatigue (β = 1.09, 95% CI = 2.12;0.01), and an indirect effect (β = 0.54, 95%CI = 1.00;0.21) via physical activity (β = 0.29, 95%CI = 0.64;0.07) and general self-efficacy (β = 0.25, 95%CI = 0.61;0.05). The intervention had a borderline significant direct effect on reduced distress (β =1.32, 95%CI =2.68;0.11), and a significant indirect effect via increased general self-efficacy and mastery (β =1.06, 95%CI = 1.89;0.38). Reductions in fatigue (β =1.33, 95%CI =1.85;0.83) and distress (β =0.86, 95%CI = 1.25;0.52) were associated with improved QoL. Further, increased physical activity was directly associated with improved QoL (β = 3.37, 95%CI = 1.01;5.54). The beneficial effect of group-based physical exercise on QoL was mediated by increased physical activity, general self-efficacy and mastery, and subsequent reductions in fatigue and distress. In addition to physical activity, future interventions should target self-efficacy and mastery. This may lead to reduced distress and fatigue, and consequently improved QoL of cancer survivors.