To examine the gastric emptying characteristics of four drinks varying in carbohydrate composition and concentration, five men ingested 600 ml of one of the different drinks on four separate occasions. All drinks contained Na+ 71 mmol.l-1, Cl- 60 mmol.l-1, Mg+2 5 mmol.l-1 and citrate 7 mmol.l-1; the carbohydrate component was either 3% glucose, 3% glucose-polymer (GP), 5% GP or 10% GP. With 99mTc-diethylene-triaminepenta-acetic acid (DTPA) as a marker, a scintillation camera and computer were used to measure the rate of gastric emptying. The half-emptying times (T 1/2) were inversely related to the glucose content of the solutions. The T 1/2 for 3% PG was 22.4 +/- 4.4 min (mean +/- SE) and for 10% GP 50 +/- 3.3 min (p less than 0.005). There was no significant difference in T 1/2 between the 3% glucose and 3% GP solutions. The increments in blood glucose (highest blood levels from 7.4 +/- 0.3 mmol.l-1 to 8.9 +/- 0.8 mmol.l-1), serum insulin (from 28 +/- 6 mU.l-1 to 77 +/- 13 mU.l-1) and C-peptide (from 3.6 +/- 0.4 micrograms.l-1 to 5.8 +/- 0.9 micrograms.l-1) were related to the amount of carbohydrate ingested. In all cases the serum insulin levels were high enough to inhibit the liberation of free fatty acids from the adipose tissue. It is concluded that the amount of carbohydrate in glucosyl units in the solution is a major determinant of gastric emptying.(ABSTRACT TRUNCATED AT 250 WORDS)