The rates of gastric emptying for highly branched cyclic dextrin (HBCD) and other carbohydrate (CHO) solutions were examined using ultrasonograph techniques. Ten healthy volunteers ingested water, physiological saline, or solutions containing various CHO, such as HBCD, glucose, maltose, sucrose, and commercially available dextrin. After a subject drank one of the solutions, the relaxed cross-sectional area of the pylorus antrum was measured at rest by real-time ultrasonography. The time required for gastric emptying was correlated with the relaxed cross-sectional area of the pylorus antrum. Among all of the solutions tested, physiological saline was transferred fastest from the stomach to the small intestine. For solutions of the same CHO, 5 % solution was transferred faster than 10 % solution. For CHO solutions other than HBCD, a low osmotic pressure was associated with rapid transfer from the stomach. The gastric emptying time (GET) of HBCD solution increased with an increase in its concentration. A shorter GET was observed for the CHO solutions at 59 to 160 mOsm regardless of their concentration. A sports drink based on 10 % HBCD adjusted to 150 mOsm by the addition of various minerals, vitamins, and organic acids was evacuated significantly (p < 0.05) faster than a 10 % HBCD solution or a sports drink based on 10 % commercially available dextrin (DE16), which has a higher osmotic pressure (269 mOsm). Our results suggest that a shorter GET could be achieved with CHO solutions with osmotic pressures of 59 - 160 mOsm. Therefore, a sports drink based on 10 % HBCD adjusted to 150 mOsm by the addition of minerals, vitamins, and organic acids could supply adequate quantities of CHO, fluid, and minerals simultaneously in a short time, without increasing GET.