In rats allowed to eat for only 2 h per day, gastric emptying of three different meals [poached egg white, glucose (0.5 kcal/ml) and physiological saline, each labelled with 60 MBq of technetium-99m tin colloid visualized using a gamma camera] was markedly slowed 8 h after the last meal. Mean body weight in rats on the restricted feeding schedule was 80% of the weight of free-feeding controls. Gastric emptying curves for all three meals in controls were best described using loge transformed counts. Other models used were linear and square root. For each of the three meals, the percentage remaining in the stomach at 120 min was estimated using linear regression of gastric contents, transformed if necessary to yield the best curve against time. At 120 min, % gastric contents (mean +/- SEM) were 19.5 +/- 5.6 (egg), 15.5 +/- 6.27 (glucose) and 27.1 +/- 7.48 (saline) in control rats. After 4 months restriction, the corresponding figures were 75.2 +/- 4.04, 81.5 +/- 4.75 and 70.3 +/- 5.83. After 3 months of free feeding, emptying rates of the three meals were not significantly different from control values. We conclude that dietary restriction causes profound changes in gastric emptying by unknown mechanisms which may be operating in patients with anorexia nervosa.