Morphine, morphine-6-glucuronide and normorphine were administered to male Sprague-Dawley-rats. Analgesic effect was estimated with the hot plate and spinal nociceptive reflex depression. After intraperitoneal administration the molar potency ratio of morphine-6-glucuronide/morphine was 1.7 estimated by the paw lick latency on the hot plate utilizing a linked pharmacokinetic-pharmacodynamic model. The potency ratio of morphine-6-glucuronide/morphine utilizing the spinal nociceptive reflex depression after intravenous administration was estimated to be within the earlier reported range of 1-4 after systemic administration of the drugs. In contrast to what is seen in man virtually no morphine-6-glucuronide was formed in Sprague-Dawley rats after administration of morphine, much lower levels of morphine-3-glucuronide were also seen. The molar AUC ratio of morphine-3-glucuronide/morphine was 1.8 +/- 0.5 and the corresponding ratio for normorphine/morphine was 0.2 +/- 0.06. After intraperitoneal administration of morphine, morphine-6-glucuronide and normorphine mean systemic clearance values of 413 +/- 95, 50 +/- 11 and 187 +/- 54 ml.min.kg-1 respectively were observed. Varea was 9.0 +/- 2.1, 0.8 +/- 0.2 and 4.9 +/- 1.4 L.kg-1 respectively. The slow absorption of morphine-6-glucuronide was illustrated by the mean Tmax-value of the 16 min. as compared with 9 min. for morphine and 10 min. for normorphine. It was possible to fit pharmacokinetic and pharmacodynamic data of behavioural analgesic effect of both morphine and morphine-6-glucuronide to a parametric model linking the sigmoid Emax model to standard pharmacokinetic equations.