The feasibility of symptom-limited cardiopulmonary exercise testing (CPET) prior to allo-SCT was assessed in addition to the prognostic value of CPET-derived measures. CPET was performed prospectively on 21 patients with hematologic malignancies, with assessments of peak (for example, peak oxygen consumption, VO2peak) and submaximal (for example, ventilatory threshold (VT)) measures of cardiopulmonary function. No serious adverse events were observed during CPET procedures, with 95% of patients achieving criteria for a peak test. Mean VO2peak was 24.7±6.4 mL kg−1min−1 (range: 10.9–35.5), equivalent to 29%±17% below that of age-matched healthy controls. All patients proceeded with the conditioning regimen followed by allo-SCT. Median follow-up was 25 months. During this period, 11 (52.4%) patients died (n = 6, relapsed disease; n = 5, non-relapse mortality (NRM)); 9 patients (43%) developed pulmonary toxicity. In univariate analyses, both peak and submaximal markers of cardiopulmonary function were predictors of OS, pulmonary toxicity and NRM. For OS, the HR for VO2peak and VT were 0.89 (95% CI, 0.8–0.99, P = 0.04) and 0.84 (95% CI, 0.71–0.98, P = 0.03), respectively. In conclusion, CPET is safe and feasible prior to allo-SCT. Patients have marked impairments in cardiopulmonary function prior to allo-SCT. CPET-derived metrics may complement conventional measures to improve risk stratification.