The purpose of this study was to find out whether the combination of inactive doses of paracetamol (PARA) and morphine was able to change dynorphin (DYN) A levels, evaluated by radioimmunoassay, and whether naloxone or [(-)-2-(3 furylmethyl)-normetazocine] (MR 2266), a kappa-opioid antagonist, modifies or prevents the activity of this combination on nociception and on DYN levels. The work was suggested by our previous findings which demonstrated that inactive doses of PARA and morphine, when given in combination, share an antinociceptive effect, and that PARA, at antinociceptive doses, decreases DYN levels in the frontal cortex, thus indicating a selective action within the CNS. Our present results demonstrate that the combination of inactive doses of PARA (100 mg/kg) and morphine (3 mg/kg) is just as effective in decreasing the levels of DYN A as full antinociceptive doses of PARA or morphine alone in the frontal cortex of the rat. The values, expressed in pmol/g tissue, were: control = 2.83 +/- 0.20; paracetamol (100) = 2.60 +/- 0.23; morphine (3) = 2.73 +/- 0.24; paracetamol + morphine = 1.34 + 0.16 (P < 0.05). The decrease was partially antagonised by MR 2266, but not by naloxone, suggesting that the activity of PARA and morphine in combination on DYN A levels could be mediated, at least in part, through kappa-receptors, although other systems may be involved. On the other hand, both naloxone and MR 2266 prevented the antinociceptive effect of the combination in the hot plate test. All our experimental data suggest that PARA and morphine in combination exert their antinociceptive effect through the opioidergic system, which in turn may cause a decrease in DYN levels in the CNS of the rat.