Radiolabeled liquid and solid portions of standardized 300-g meals were administered on four different study days to eight healthy subjects in an attempt to define the range of inter- and intrasubject variability in gastric emptying. Meal half emptying times, analysis of variance, and intraclass correlations were computed and compared within and between subjects. The mean solid half emptying time was 58 +/- 17 min (range 29-92), while the mean liquid half emptying time was 24 +/- 8 min (range 12-37). A nested random effects analysis of variance showed moderate intrasubject variability for solid emptying and high intrasubject variability for liquid emptying. The variability of solid and liquid emptying was comparable and relatively large when compared with other reports in the literature. The isotopic method for measuring gastric emptying is a valuable tool for investigating problems in gastric pathophysiology, particularly when differences between groups of subjects are sought. However, meal emptying time is a variable phenomenon in healthy subjects with significant inter- and intraindividual day-to-day differences. These day-to-day variations in gastric emptying must be considered in interpreting individual study results.