The effects of acute, systemic administration of amitriptyline, duloxetine and mirtazapine (antidepressant drugs that variously affect extracellular noradrenaline and serotonin levels) and the selective serotonin reuptake inhibitor (SSRI) citalopram were compared in rat models of experimental pain. None of the drugs (all 3-30 mg/kg, i.p.) affected acute nociceptive responses as measured in the tail flick test. In the hot plate test, duloxetine and mirtazapine significantly increased (P<0.05) the nociceptive response latency, whereas amitriptyline and citalopram were ineffective. In the formalin test, duloxetine and citalopram significantly attenuated, whereas amitriptyline and mirtazapine increased, second phase flinching behaviour (all P<0.05). However, amitriptyline and mirtazapine reduced second phase licking behaviour. In the chronic constriction injury model of neuropathic pain, thermal hyperalgesia of the injured hindpaw was significantly attenuated by all four drugs (P<0.05); only amitriptyline and duloxetine fully reversed thermal hypersensitivity. None of the drugs tested attenuated mechanical allodynia. In contrast amitriptyline, duloxetine and mirtazapine significantly reduced mechanical hyperalgesia (P<0.05); citalopram was ineffective. No drug-related effects on motor performance in the rotarod test were observed. These results (a) highlight the difficulty in correlating antinociceptive effects of drugs from different antidepressant classes across a range of animal pain models and (b) suggest that antidepressants that variously affect both noradrenaline and serotonin levels have more potent and efficacious antinociceptive effects than SSRIs (as exemplified by citalopram), against a range of pain-like behaviours in an animal model of neuropathic pain.