The streptozocin-induced diabetic rat has been put forward as a model of chronic pain with signs of hyperalgesia and allodynia that may reflect signs observed in diabetic humans. The aim of this work was to assess, in streptozocin-induced diabetic rats, the pharmacological activity to several analgesic drugs known to be effective (clomipramine, amitriptyline, desipramine, clonidine, lidocaine), ineffective (aspirin), or with a doubtful effectiveness (morphine) in human painful diabetic neuropathy. The animals were submitted to a mechanical pain test (paw pressure) and the ability of the drugs to reverse diabetes-induced hyperalgesia was tested. The tested antidepressants (0.125-8 mg/kg, i.v.) were slightly effective in diabetic rats; amitriptyline and clomipramine induced a weak effect, whereas desipramine was more active, suggesting noradrenergic specificity. This was confirmed by the effectiveness of clonidine (50, 100, 150 micrograms/kg, s.c.). Lidocaine (1-9 mg/kg, i.v.) had prolonged efficacy on mechanical hyperalgesia. Aspirin (100 mg/kg, i.v.) was without effect and morphine (0.5-4 mg/kg, i.v.) induced a dose-dependent antinociceptive effect but at doses twice as high as those used in normal rats. These results demonstrate the high pharmacological predictivity of this model of painful diabetes and suggest that in this pathological condition, among the drugs acting on monoaminergic transmission, noradrenergic drugs seem the most active.