In the decerebrated rabbit intravenous fentanyl, a μ-opioid agonist, acts at spinal and supraspinal sites to inhibit reflex responses in medial gastrocnemius (MG) motoneurones evoked by electrical stimulation of the sural nerve. This inhibition is supported by descending adrenergic pathways thus spinal section or intrathecal (i.th.) administration of the selective α₂-adrenoceptor antagonist RX 821002 reduces the effect of fentanyl (Clarke et al., 1998). Studies with i.th. morphine however suggest that this opioid does not interact with endogenous noradrenaline in the same way (Clarke & Lo, 2004). The present study has examined this relationship further using i.v. morphine. Experiments were performed on 26 rabbits decerebrated under isoflurane (3-5 %)/N₂O anaesthesia. Eight of these animals were also spinalized at L1. Electrical stimulation of the left sural nerve at C fibre intensity (144 times threshold) was used to evoke reflex responses in the ipsilateral MG muscle nerve. These were averaged, integrated by computer and analyzed in 3 post-stimulus time bands: 5-12 ms (phase 1); 12-100 ms (phase 2) and 100-250 ms (phase 3). After a control period of at least 30 min, i.v. morphine was injected at 30 min intervals in doses of 1, 2, 7 and 20 mg kg^-1 to give a total cumulative dose of 30 mg kg^-1. In 9 non-spinalized animals RX 821002 (100 μg i.th.) was given prior to i.v. morphine. Experiments were terminated by i.v. injection of saturated KCl solution. In decerebrated animals, i.v. morphine significantly (Friedman′s ANOVA, p < 0.0001, n = 9) inhibited phase 1, 2 and 3 reflexes so that after 30 mg kg^-1 cumulative, median responses were 7%, 15% and 11% of pre-drug levels respectively. In animals pre-treated with RX 821002, i.v. morphine had no effect on phase 3 responses but it did significantly (Friedman′s ANOVA, p < 0.002, n = 9) inhibit phase 1 and 2 reflexes to 29% and 39% of pre-morphine controls after 30 mg kg^-1: these values were not significantly different to those in the absence of RX 821002 (Mann-Whitney tests, p > 0.2). In the spinalized group, responses were significantly (Friedman′s ANOVA, p < 0.004, n = 8) inhibited by morphine to 58%, 31% and 20% of control values for phases 1, 2 and 3 respectively. For phase 1 reflexes this inhibition was significantly reduced compared to spinally intact animals (Mann-Whitney test, p < 0.02). These data corroborate our findings using i.th. morphine and show that inhibition of short-latency reflexes by i.v. morphine is partially mediated by descending pathways. However, adrenergic systems appear to support morphine only against long-latency reflexes.