To determine whether increases in muscle mitochondrial capacity are necessary for the characteristic lower exercise glycogen loss and lactate concentration observed during exercise in the trained state, we have employed a short-term training model involving 2 h of cycling per day at 67% maximal O2 uptake (VO2max) for 5-7 consecutive days. Before and after training, biopsies were extracted from the vastus lateralis of nine male subjects during a continuous exercise challenge consisting of 30 min of work at 67% VO2max followed by 30 min at 76% VO2max. Analysis of samples at 0, 15, 20, and 60 min indicated a pronounced reduction (P less than 0.05) in glycogen utilization after training. Reductions in glycogen utilization were accompanied by reductions (P less than 0.05) in muscle lactate concentration (mmol/kg dry wt) at 15 min [37.4 +/- 9.3 (SE) vs. 20.2 +/- 5.3], 30 min (30.5 +/- 6.9 vs. 17.6 +/- 3.8), and 60 min (26.5 +/- 5.8 vs. 17.8 +/- 3.5) of exercise. Maximal aerobic power, VO2max (l/min) was unaffected by the training (3.99 +/- 0.21 vs. 4.05 +/- 0.26). Measurements of maximal activities of enzymes representative of the citric acid cycle (succinic dehydrogenase and citrate synthase) were similar before and after the training. It is concluded that, in the voluntary exercising human, altered metabolic events are an early adaptive response to training and need not be accompanied by changes in muscle mitochondrial capacity.