Steady state pharmacokinetics of morphine (M), morphine-3-glucuronide (M3G) and morphine-6-glucuronide (M6G) were investigated in 6 patients with intractable cancer pain administered orally with MST (Mundipharma, Limburg, Germany) and, subsequently, rectally with MSR to make a judgment whether orally administered morphine can be replaced by rectally administered morphine. The parent drug and glucuronide metabolites were measured simultaneously using high-performance liquid chromatography (HPLC) and native fluorescence detection. The mean morphine area under the curve (AUC) value (0-8 h) was smallest (434.3 ± 170.2 nmolL-1h) in the oral administration than in the rectal administration (574.8 ± 285.0 nmolL-1h) (p < 0.05). The rectal administration resulted in less production of M3G and M6G. There were no significant differences in the mean steady state concentrations (Css) of morphine, M3G, and M6G between the oral and rectal administrations (p > 0.05). The median AUC ratio-M3G/M and M6G/M, 12.58 and 1.85-following MSR rectal administration was smaller than following MST oral administration in 6 patients (19.97 and 2.59; p < 0.05), whereas the median AUC ratio M3G/M6G in the rectal dosing was 6.24 (range 5.2-7.6) was almost the same as the median ratio M3G/M6G in the oral dosing was 6.49 (range 5.8-8.5; p > 0.1). Four of the 6 patients had a greater Cmax of M3G and M6G after oral administration than after rectal administration. The same 4 had lower fluctuation rates for morphine. M3G (p < 0.05), and M6G (p < 0.05) after rectal administration. Therefore, during chronic morphine treatment, it still seems difficult to decide whether oral administration can be replaced by rectal administration.