Antinociceptive tolerance to morphine (MOR) was induced in groups of Sprague-Dawley rats receiving continuous intravenous infusions of morphine sulphate administered by 3 different MOR dosing regimes. At appropriate intervals throughout each infusion period, antinociceptive testing was performed using the tail-flick latency test and blood samples were collected. Groups of saline (SAL)-infused control rats also underwent antinociceptive testing and blood sample collection. Complete antinociceptive tolerance developed during each MOR infusion period and was characterized by a marked decline in the degree of antinociception from values greater than 90% of the maximum possible effect (%MPE) to pre-dosing baseline values. By contrast, %MPE values in SAL-infused control animals and in sham-operated rats were not significantly different from pre-dosing values throughout the infusion period, indicating that the experimental procedures themselves did not contribute to the development of antinociceptive tolerance to MOR. In addition, the rate of MOR tolerance development was inversely proportional to the MOR infusion rate. A very significant inverse relationship was observed between the mean degree of antinociception (%MPE) and the mean plasma molar concentration ratio, [morphine-3-glucuronide]/[MOR], for each of the 3 MOR dosing regimes and for the cumulated data. This relationship showed that near-maximum antinociception was attainable at ratio values less than approximately 0.50, whilst at ratio values above approximately 1.5, little or no antinociception was observed. Although %MPE was highly inversely correlated with the mean plasma morphine-3-glucuronide (M3G) concentrations for rats receiving regimes A and B, this was not the case for rats receiving regime C where antinociceptive tolerance was partially reversed by an increase in the morphine infusion rate part-way through the infusion period. In addition, a poor relationship was observed between %MPE and the mean plasma MOR concentration, possibly due to the confounding presence of M3G in all samples. Thus, we may conclude from this study in Sprague-Dawley rats that irrespective of the rate of antinociceptive tolerance development, the level of antinociception achievable appears to be highly inversely correlated with the mean [M3G]/[MOR] plasma molar concentration ratio and poorly correlated with the plasma MOR concentration, consistent with the notion that it is perhaps the balance between the excitatory effects of M3G and the inhibitory effects of MOR at the functional level which is the important determinant. Further research is required in carefully conducted studies in cancer patients to evaluate the possible contribution of the MOR metabolites, M3G and morphine-6-glucuronide (MbG), to increasing dosing requirements of MOR.(ABSTRACT TRUNCATED AT 400 WORDS)