The bioavailability of morphine after rectal administration using three types of suppositories containing morphine hydrochloride (10 mg) in different added forms was evaluated in rabbits. Three types of suppositories were constructed with a base material (Witepsol H-15): a conventional suppository containing morphine hydrochloride mixed with a base material, a hollow-type suppository containing morphine in an aqueous solution in its cavity, and a hollow-type suppository containing morphine as a powder (hydrochloride salt) in its cavity. The plasma concentrations of morphine and its metabolites, morphine-3-glucuronide (M-3-G) and morphine-6-glucuronide (M-6-G), were determined. The mean AUC0-6 of morphine after rectal administration of the hollow-type suppository containing powdered morphine was significantly higher than that after the administration of a conventional suppository, whereas the mean AUC0-6 of M-3-G was lower than that after administration of a conventional suppository. The mean AUC0-6 of M-6-G by a hollow-type suppository containing powdered morphine was higher than that by a conventional suppository. Although the mean AUC0-6 ratio of M-3-G to morphine after administration of a conventional suppository was three times larger than that of a hollow-type suppository containing powdered morphine, the mean AUC0-6 ratios of M-6-G to morphine after use of the three types of suppositories were all approximately 1.0. This study demonstrates that the hollow-type suppository containing powdered morphine is a more effective rectal dosage vehicle than the conventional suppository.