Using extracellular single-unit recording techniques, effects of intravenously administered lidocaine on dorsal-horn nociceptive neurons were studied in cats made decerebrate whose spinal cords had been transected. Thirty-seven neurons in Rexed lamina V responding to high-threshold mechanical and noxious thermal stimuli (radiant heat, using Hardy-Wolff-Goodell dolorimeter) were studied. Lidocaine hydrochloride, 2.5, 5, and 10 mg/kg, iv, produced dose-related suppression of both spontaneous activity and responses of these neurons to noxious thermal stimulation. Spontaneous discharge frequencies at maximum suppression, observed 3–7 min after administration of each of the three doses of lidocaine were 64 ± 14 (mean · 1 SE), 32 ± 8, and 25 ± 9 per cent of control values, respectively; responses to noxious thermal stimuli were 83 ± 5, 52 ± 8, and 39 ± 7 per cent of the control values, respectively. Threshold skin temperature to noxious thermal stimulation increased from 44.7 ± 0.4 C (control) to 46.3 ± 0.7 C with lidocaine, 5 mg/kg (P < 0.05), to 47.8 ± 0.8 C with lidocaine, 10 mg/kg (P < 0.01). The times necessary for recovery varied in a dose-related fashion. Plasma lidocaine concentrations 5 min after lidocaine, 5 mg/kg, averaged 3.6 ± 0.7 μg/ml. These data support the clinical impression that intravenously administered lidocaine produces analgesia at plasma concentrations of 3–10 μg/ml. It is suggested that lidocaine may block conduction of nociceptive impulses, at least in part, by suppression of spinal-cord nociceptive neurons.