The effect of a 3-hour versus a 10-hour preoperative fasting interval on the gastric residual volume and gastric pH of pediatric patients was evaluated. Forty-four healthy infants, 1 month to 5 years of age, were randomly assigned to one of two groups. The 3-hour nil per os (NPO) group consisted of 19 infants kept NPO for 3 hours following ingestion of up to 4 ounces of 5% dextrose in water (D5W). The control group consisted of 25 infants who remained NPO an average of 10 hours prior to surgery. Gastric residual volume was calculated using the dye-dilution technique. After the dye marker was injected into the stomach, complete aspiration of the stomach (including the volume of dye marker plus residual gastric contents) was attempted as another method to measure gastric residual volume. There were no significant differences in gastric residual volume between the 3-hour and the 10-hour NPO groups using either the dye-dilution or aspiration methods. However, there were significant differences between the two measuring techniques. Gastric residual volume was significantly greater in volume when measured by the dye-dilution technique than it was when measured by the aspiration technique in both the 10-hour (p less than 0.009) and the 3-hour (p less than 0.0009) NPO groups. Complete aspiration of a known volume of fluid injected through the orogastric tube was not possible in 23 of the 44 (52.4%) infants. Mean gastric pH was less than 2.0 in both groups.(ABSTRACT TRUNCATED AT 250 WORDS)