BACKGROUND: The investigation of exercise training in metastatic breast cancer has received minimal attention. This study determined the feasibility and safety of aerobic training in metastatic breast cancer. METHODS: Sixty-five women (age, 21-80 years) with metastatic (stage IV) breast cancer (57% were receiving chemotherapy, and >40% had ≥ 2 lines of prior therapy) were allocated to an aerobic training group (n = 33) or a stretching group (n = 32). Aerobic training consisted of 36 supervised treadmill walking sessions delivered thrice weekly between 55% and 80% of peak oxygen consumption (VO2peak ) for 12 consecutive weeks. Stretching was matched to aerobic training with respect to location, frequency, duration, and intervention length. The primary endpoint was aerobic training feasibility, which was a priori defined as the lost to follow-up (LTF) rate (<20%) and attendance (≥70%). Secondary endpoints were safety, objective outcomes (VO2peak and functional capacity), and patient-reported outcomes (PROs; quality of life). RESULTS: One of the 33 patients (3%) receiving aerobic training was LTF, whereas the mean attendance rate was 63% ± 30%. The rates of permanent discontinuation and dose modification were 27% and 49%, respectively. Intention-to-treat analyses indicated improvements in PROs, which favored the attention control group (P values > .05). Per protocol analyses indicated that 14 of 33 patients (42%) receiving aerobic training had acceptable tolerability (relative dose intensity ≥ 70%), and this led to improvements in VO2peak and functional capacity (P values < .05). CONCLUSIONS: Aerobic training at the dose and schedule tested is safe but not feasible for a significant proportion of patients with metastatic breast cancer. The acceptable feasibility and promising benefit for select patients warrant further evaluation in a dose-finding phase 1/2 study. Cancer 2018;124:2552-60. © 2018 American Cancer Society.