The postexercise recovery period is characterized by transition from the acutely catabolic phase that occurs during exercise to an anabolic phase. Muscle glycogen stores are replenished after exercise and even in the absence of refeeding have returned to 50% of preexercise concentrations within 12 h (1). The increase in net protein degradation that is demonstrable during exercise at low-to-moderate workboads is replaced by net protein resynthesis as soon as exercise is stopped (2, 3). These anabobic responses are essential for physical training to result in increased glycogen stores and lean body mass. Physical training programs have been advocated as an important component in the treatment of subjects with noninsulin-dependent diabetes mellitus (NIDDM). Physical training was shown to improve the major defect present in NIDDM, ie, peripheral insulin resistance (4), and to improve glycemic control in these subjects (5). How much ofthis beneficial effect ofregular physical exercise is due to the cumulative effects of single bouts of exercise vs a specific training effect has been unclear. Recent work suggests that benefits may be demonstrable after single bouts ofexercise (6, 7) and the training effect is rapidly lost when an exercise program is interrupted (7, 8). Although most exercise studies have focused on effects on peripheral (skeletal muscle) insulin sensitivity, hepatic glucose production may also be affected by exercise(6). Although postexercise recovery is characterized by overall anabobic responses, these responses may vary in different body compartments. Ahlborg et al (9) showed that nonexercised muscle demonstrates ongoing catabolism (glycogenolysis) during the early postexercise period (40 mm) with release ofbactate presumably for gluconeogenesis. Whether this differential response (catabolism in nonexercised muscle, anabolism in exercised muscle) occurs for longer periods and involves other gluconeogenic precursors requires further study. Finally, food ingestion during exercise was shown to enhance endurance exercise performance in some studies ( 10) but not in others ( 11). Preexercise glucose ingestion may adversely affect performance (1 2). We (13) studied the effects of preexercise ingestion of a mixedmeal snack on the metabolic fuel responses during and after exercise. Subjects for the studies conducted during the bate (1216 h) postexercise recovery period consisted of six male control subjects within 18% of ideal body weight (14), designated as lean for the purpose of the study; six male subjects > 130% ideal body weight with either glucose intolerance or hyperinsubinemia during a 75-g oral glucose tolerance test ( 15), designated as obese; and five male subjects with NIDDM (1 5) who were all > 1 18% ideal body weight. Subjects for the studies conducted during early (2-4 h) postexercise recovery consisted of 1 1 lean healthy subjects (6 male, 5 female). For the studies of the effects of preexercise snack feeding, nine healthy male subjects within 15% of ideal body weight were selected. Except for obesity and diabetes melbitus in the respective groups, all subjects were deemed to be in excellent health as determined by medical history, physicab examination, and laboratory screening chemistry studies.