Although peak oxygen consumption (VO(2)) during exercise is frequently used to help predict optimal timing for heart transplantation, its long-term prognostic value in women is not known. We followed 2,105 adult patients with heart failure (HF) and with impaired left ventricular (LV) systolic function for 5 years, including 525 women (25%) who underwent metabolic stress testing between January 1995 and December 2002. Multivariable proportional hazards modeling related VO(2) to survival with adjustments for >30 confounders and with transplantation considered as a time-dependent covariate. During follow-up, 129 women (26%) died, as did 572 men (36%). There were 175 transplants, including 34 women. Women and men were similar in age (54 vs 55 years), but women were less likely to have coronary artery disease (28% vs 58%). Peak VO(2) was strongly predictive of time to death in women (adjusted hazard ratio [HR] for peak VO(2) decreasing from 15 to 14 ml/kg/min, 1.11, 95% confidence interval [CI] 1.05 to 1.18, p <0.0001) and in men (adjusted HR 1.12, 95% CI 1.08 to 1.16, p <0.0001). There was no gender interaction with peak VO(2) (p = 0.80), but for any given peak VO(2) women were at lower risk (adjusted HR for men compared with women 2.22, 95% CI 1.58 to 3.10, p <0.0001). A significant interaction was found between gender and presence of coronary artery disease (p for interaction 0.02); in women, those with ischemic cardiomyopathy had a worse survival for any given peak VO(2). In conclusion, in this large cohort, peak VO(2) predicted survival in women and men whether or not coronary artery disease was present, but an interaction was noted between coronary artery, gender, and survival.