The rate of muscle glycogen synthesis during 2 and 4 h of recovery after depletion by exercise was studied using two energy equivalent carbohydrate drinks, one containing a polyglucoside with a mean molecular mass of 500 000-700 000 (C drink), and one containing monomers and oligomers of glucose with a mean molecular mass of approximately 500 (G drink). The osmolality was 84 and 350 mosmol. l(-1), respectively. A group of 13 healthy well-trained men ingested the drinks after glycogen depleting exercise, one drink at each test occasion. The total amount of carbohydrates consumed was 300 g (4.2 g. kg(-1)) body mass given as 75 g in 500 ml water immediately after exercise and again 30, 60 ad 90-min post exercise. Blood glucose and insulin concentrations were recorded at rest and every 30 min throughout the 4-h recovery period. Muscle biopsies were obtained at the end of exercise and after 2 and 4 h of recovery. Mean muscle glycogen contents after exercise were 52.9 (SD 27.4) mmol glycosyl units. kg(-1) (dry mass) in the C group and 58.3 (SD 35.4) mmol glycosyl units. kg(-1) (dry mass) in the G group. Mean glycogen synthesis rate was significantly higher during the initial 2 h for the C drink compared to the G drink: 50.2 (SD 13.7) mmol. kg(-1) (dry mass). h(-1) in the C group and 29.9 (SD 12.5) mmol. kg(-1) (dry mass). h(-1) in the G group. During the last 2 h the mean synthesis rate was 18.8 (SD 33.3) and 23.3 (SD 22.4) mmol. kg(-1) (dry mass). h(-1) in the C and G group, respectively (n.s.). Mean blood glucose and insulin concentrations did not differ between the two drinks. Our data indicted that the osmolality of the carbohydrate drink may influence the rate of resynthesis of glycogen in muscle after its depletion by exercise.