In an attempt to measure gastric emptying of carbohydrate solutions after exercise, we used the 13C acetate breath test to differentiate the gastric emptying of three approximately isoenergetic carbohydrate solutions (i.e. glucose, glucose polymer and sucrose) from each other and from water. On four separate occasions, six post-absorptive subjects walked on an inclined treadmill at 70% maximum oxygen uptake for 1 h and were then given 330 ml of one of the solutions in which 150 mg of sodium 1-[13C] acetate had been dissolved. Breath samples were collected at regular (2-30 min) intervals over the next 3.5 h for analysis of expired 13CO2 by isotope ratio mass spectrometry. When water was given, all subjects reached peak breath enrichment after 30 min, and had a mean (SE) gastric emptying time of 33.2 (1.6) min. Peak breath enrichment occurred later for sucrose and glucose polymer at 54.3 (3.1) min and 59.0 (2.1) min respectively (P < 0.01), and for glucose this was even later, at 62.3 (1.0) min (P < 0.05). Calculated gastric emptying times for sucrose and glucose polymer were almost identical [66.5 (2.5) and 69.8 (2.9) min respectively], whereas that for glucose was significantly slower [76.8 (3.2) min; P < 0.02], probably reflecting the effects of increased osmolality. The gastric emptying of all carbohydrates were significantly longer than for water (P < 0.01). These results show that in the post-exercise state the 13C acetate breath test can be used to differentiate the gastric emptying rates of water and carbohydrate solutions of different properties.