Twenty-six patients with severe pain associated with cancer were entered into a study where they were required to take morphine mixture for 7 days. Prior to this, their morphine dose had been optimised to provide the most favourable balance between pain relief and side effects. After 6 days of taking their optimised morphine dose at 4-hourly intervals, the patients were admitted to the Pain Management Unit such that the doses from 18:00 h on day 6 were taken under direct nursing supervision. Frequent blood samples were collected after the 10:00 h (dose 1), 14:00 h (dose 2) and 18:00 h (dose 3) on day 7. There was a significant difference between the 3 doses with respect to Cmax values for morphine with dose 3 > dose 1 > dose 2. Further, there was considerable variability in the percentage change of either dose 1 or dose 3 Cmax values relative to dose 2. The Cmax values of the active metabolite morphine-6-glucuronide (M6G), measured in 6 of the patients, for the 3 dosing intervals followed a similar trend to the parent drug, but only doses 1 and 2 differed significantly. Similar but less pronounced changes were observed in the area-under-curve parameter calculated for both morphine and M6G during the 3 dosing intervals. There were no significant differences in the Cmin or Tmax parameters for either morphine or M6G between the 3 dosing intervals. These results suggest intra-individual variation in the absorption of morphine or changes in the volume of distribution during the day.(ABSTRACT TRUNCATED AT 250 WORDS)