The aim of this study was to assess the nature and magnitude of the differences in submaximal and maximal exercise capacity parameters between lean and obese women. A total of 225 healthy obese women 18–65 years (BMI> or=30 kg/m2) and 81 non-athletic lean women (BMI< or=26 kg/m2) were selected. Anthropometric measurements (weight and height), body composition assessment (bioelectrical impedance method) and a maximal exercise capacity test on a bicycle ergometer were performed. Oxygen uptake (VO2), carbon dioxide production (VCO2), expired ventilation (VE), respiratory quotient (RQ), breathing efficiency (VE/VO2), mechanical efficiency (ME) and anaerobic threshold (AT) were calculated. At a submaximal intensity load of 70 W, VO2 (l/min) was larger in the obese women and was already 78% of their peak VO2, whereas in the non-obese it was only 69% (P=0.0001). VE (l/min) was larger, VE/VO2 did not differ and ME was lower in obese compared to the lean women. AT occurred at the same percentage of peak VO2 in both lean and obese women. At peak effort, achieved load, terminal VO2 (l min−1 kg−1), VE, heart rate, RQ respiratory exchange ratio and perceived exertion were lower in obese subjects compared to the non-obese. Obese subjects mentioned significantly more musculoskeletal pain as a reason to end the test, whereas in lean subjects it was leg fatigue. Lean women recovered better as after 2 min they were already at 35% of the peak VO2, whereas in the obese women it was 47% (P=0.0001). Our results confirm that exercise capacity is decreased in obesity, both at submaximal and peak intensity, and during recovery. Moreover, at peak effort musculoskeletal pain was an important reason to end the test and not true leg fatigue. These findings are important when designing exercise programs for obese subjects.