Energy restriction (ER) has shown to be an effective 'anti-aging' factor in rodents, resulting in an increased life span and preventing or delaying the occurrence of many age-related diseases in rodents. As a part of a feasibility study on the potential application of ER in humans, we studied the effects of moderate ER on physical performance in 24 apparently healthy, non-obese middle-aged men. After two weeks of weight maintenance the ER group (n = 16) received 80% of their habitual energy intake, while a control group (n = g) still received their weight-maintaining diet for ten weeks. Physical performance (bicycle ergometer) was estimated by a maximal (until exhaustion) and a submaximal (30min at 60% of VO2max) exercise test. After the experimental period the subjects in ER group had a significantly shorter cycling time (Δ = 1.31 ± 1.14 vs - 0.45 ± 1.56 min, p = 0.01), and thus had a lower maximal power output (Δ = 4.3 ± 5.9 vs -6.3 ± 13.6 watt, p < 0.01) and maximal oxygen uptake (Δ = 0.03 ± 0.18 vs - 0.18 ± 0.32 1/min, p = 0.05) during the maximal exercise test, than the subjects in the control group. None of these changes correlated with weight loss. As in every exercise test, motivation could have played an important role. During the submaximal exercise test no significant changes between the two groups in HR, VO2, RER and FFA, glucose, lactate and catecholamine levels were observed. It is concluded that 10 weeks of moderate ER lead to a slight but significant decrease in maximal power output, which is not related to weight loss. Moreover, no changes on submaximal physical performance and on substrate metabolism have been observed although significant alterations in body weight (-7.4 ± 2.6 kg) occurred.