Abstract Background Exercise training (ExT) is emerging as a useful therapy to address limitations in VO2peak in anthracycline-treated breast cancer (BC) survivors. However, the factors necessary for improving VO2peak during (neo)adjuvant BC therapy remain unclear. Purpose To explore the physiologic factors that were associated with a meaningful increase in VO2peak following a 12-month structured ExT intervention in breast cancer survivors undergoing anthracycline-based therapy. Methods This was a secondary analysis of the BReast cancer EXercise InTervention (BREXIT) Trial using baseline and 12-month data from participants randomized to ExT with complete VO2peak data at baseline and 12-months (n=48). A meaningful increase in VO2peak was defined as a ≥1.65 mL/kg/min improvement in VO2peak from baseline (pre-chemotherapy) to 12-months (n=22, 46% of ExT group) as this corresponds to the minimal detectable change from our laboratory and exceeds previously published values for a minimally clinically important difference. VO2peak, peak heart rate (HRpeak) and peak arterial O2 saturation were assessed from a maximal cardiopulmonary exercise test. Peak exercise cardiac function (stroke volume [SVpeak] and cardiac output [Qcpeak]) was assessed by exercise cardiac magnetic resonance. Peak exercise convective O2 delivery (QO2peak), arterio-venous oxygen difference (a-vO2diff) and skeletal muscle diffusive O2 conductance (O2 transport from microvasculature to mitochondria [DMO2]) were calculated from VO2peak, arterial O2 saturation, exercise hemodynamic measures and haemoglobin concentration (assessed via venous blood draw) according to previously published equations. Results There were no significant differences in age, body mass index, baseline VO2peak or cardiac function between those who increased VO2peak or not. Those with increased VO2peak had higher ExT adherence (78% vs 68%, P=0.005). As expected, the increased VO2peak group had a 21% improvement in VO2peak from baseline to 12-months compared to the no change in the no VO2peak increase group (-2%; interaction P<0.001). Both groups showed comparable increases in SVpeak (+12% vs +7%, interaction P=0.09), Qcpeak (+13% vs +11%, Interaction P=0.48) and QO2peak (+14% vs +7%, interaction P=0.13). The key factors differentiating the two groups was that the increased VO2peak group also increased DMO2 (+17%, interaction P<0.001), that acted to maintain the a-vO2diff. In contrast, the group with stable VO2peak experienced no change in DMO2, which resulted in a decrease in the a-vO2diff (-12%, interaction P<0.001) and explains why they saw negligible change in VO2peak. Conclusion In anthracycline-treated BC survivors, achieving a meaningful improvement in VO2peak with exercise training requires concurrent adaptations in factors governing O2 delivery (peak cardiac function) and diffusion of O2 into skeletal muscle, highlighting the importance of multi-modal exercise training interventions in this population.