The aim of this study was to determine the effects of olestra and sorbitol consumption on three accepted objective measures of diarrhea (stool output >250 g/day, liquid/watery stools, bowel movement frequency >3/day), and how stool composition influences reports of common gastrointestinal symptoms. A double-blind, placebo-controlled study compared the effects of sorbitol (40 g/day in candy), a poorly absorbed sugar-alcohol with known osmotic effects, with those of olestra (20 or 40 g/day in potato chips), a nonabsorbed fat, on objective measures of stool composition and GI symptoms. Sixty-six subjects resided on a metabolic ward for 12 days: 2 days lead-in, 4 days baseline, 6 days treatment. Sorbitol 40 g/day resulted in loose/liquid stools within 1-3 h of consumption. In contrast, olestra resulted in a dose-responsive stool softening effect after 2-4 days of consumption. Subjects reported "diarrhea" when mean stool apparent viscosity (peak force (PF), g) decreased from a perceived "normal" (mean +/- SE, 1355 +/- 224 g PF; firm stool) to loose (260 +/- 68 g PF) stool. Mean apparent viscosity of stool during treatment: placebo, 1363 +/- 280 g (firm); olestra 20 g/day 743 +/- 65 g (soft); olestra 40 g/day, 563 +/- 105 g (soft); and sorbitol 40 g/day, 249 +/- 53 g (loose). Of the 1098 stool samples collected, 38% (419/1098) were rated by subjects as "diarrhea," yet only 2% of treatment days (all in the sorbitol treatment group) met commonly accepted criteria for a clinical diarrhea. Sorbitol, but not olestra, increased the severity of abdominal cramping, urgency and nausea compared to placebo. Olestra consumption, at levels far in excess of normal snacking conditions, resulted in a gradual stool softening effect after several days of consumption, did not meet any of the three objective measures of diarrhea, and did not increase GI symptoms. Sorbitol consumption, at only 80% of the dose requiring a "laxative effect" information label, resulted in rapid onset loose/liquid stools and a significant increase in abdominal cramping, urgency and nausea. Overall, subjects categorized stool as "diarrhea" when stool decreased from their perceived "normal," but the vast majority of these reports were not associated with clinically significant diarrhea.