To investigate further the hormonal and metabolic adaptations occurring when carbohydrates are ingested after prolonged exercise, we have compared the fate of a 100-g oral glucose load (using 'naturally labelled' 13C-glucose) in healthy volunteers after an overnight fast at rest either without previous exercise or after a 3-h exercise performed on a treadmill at about 50% of the individual VO2 max. In comparison to the control conditions, the oral glucose tolerance test (OGTT) performed in the post-exercise recovery period was characterized by a greater rise in peripheral blood glucose levels and delayed insulin response. Plasma glucagon values were significantly elevated at the time glucose was given (+48 +/- 13 pg ml-1) and at the end of the OGTT. Plasma-free fatty acid (FFA) levels were 1675 +/- 103 microEq 1-1 when glucose was given, and subsequently reduced to values similar to those observed in the control conditions. Indirect calorimetry indicated that OGTT in post-exercise recovery was associated with decreased carbohydrate and increased lipid oxidation when compared to control conditions. Exogenous glucose oxidation was also significantly reduced: 21.1 +/- 2.6 vs. 35.9 +/- 1.9 g per 7 h. We suggest that the higher plasma glucagon levels and the delayed insulin response played a role in the decreased hepatic glucose retention previously described by others in post-exercise recovery. Our data also suggest that the higher lipid oxidation rate observed at the time glucose was given in the post-exercise period could explain, according to the Randle 'glucose-fatty acid cycle', the decreased carbohydrate oxidation and the preferential muscle glycogen repletion already well documented. The reason why the lipid oxidation rate remains increased 3-7 h after glucose ingestion in spite of the fact that FAA levels at that time are similar to those observed in control conditions is still unknown; further kinetic studies are needed to clarify this point.