The suitability of gastric-emptying-controlled rabbits (GEC-rabbits) as a model for estimating the bioavailability of controlled-release preparations in humans was investigated. Non-disintegrating tablets (diameter of 4.1-7.7 mm) and granules (diameter of 0.8-2.0 mm ; specific gravity of 0.90-1.85) were used as a model preparation. Six tablets and 100 granules were administered to GEC-rabbits, and the number of each dosage form remaining in the gastrointestinal tract was counted at suitable intervals. Tablets with diameters of 4.1 and 5.8 mm were randomly emptied from the stomach and the inter-animal variation was very large. Tablets with a diameter of 7.7 mm were not emptied at all even after 7 h. Granules were gradually emptied from the stomach and the profile was similar to that in humans under non-fasting conditions. The gastric emptying was strongly influenced by specific gravity, but the small intestinal transit time was not influenced by either diameter or specific gravity. The mean transit time through the small intestine (about 1.2 h) was about one-third that of humans. The results obtained in this study show that it is important to consider duration and extent of absorption in attempting to predict the bioavailability of controlled-release preparations in humans by using GEC-rabbits.