1. The effect of chlormethiazole has been studied in a transient middle cerebral artery (MCA) occlusion model of cerebral ischaemia in the rat. The MCA was occluded for 1 h by use of an intraluminal suture technique, with reperfusion for 24 h following removal of the occluding filament. Neuronal damage was determined by measurement of the area of necrosis following Cresyl Violet staining of sections taken through the ischaemic region. 2. In the initial experiment, occlusion of the MCA produced a large volume of ischaemic damage in both cortex and striatum, characterized by necrosis and pyknosis (total volume of damage, 287 +/- 13 mm3, n = 9). Rats injected with chlormethiazole (1000 mumol kg-1, i.p.) 60 min before occlusion had a reduced volume of damage in both regions (104 +/- 11 mm3; n = 9; P < 0.001). 3. In a subsequent study systemic physiological parameters (heart rate, blood pressure, blood pH, blood gases and rectal temperature) were measured throughout the ischaemic period. 4. Chlormethiazole (1000 mumol kg-1) pretreatment produced little change in systemic physiology and the neuroprotective effect of the drug when given 60 min prior to the MCA occlusion was confirmed. Chlormethiazole was also neuroprotective when given 10 min following the start of reperfusion (control group: 244 +/- 52 mm3, n = 10; chlormethiazole pretreatment group: 102 +/- 23 mm3, n = 10; P < 0.001; chlormethiazole post-ischaemia group: 122 +/- 16 mm3; P < 0.001, n = 10). 5. It is concluded that chlormethiazole is an effective neuroprotective agent in this model of transient focal ischaemia. The observation that chlormethiazole is protective when given after reperfusion indicates that the effect of the drug is unlikely to be due to an alteration of intra-ischaemic cerebral blood flow, but is more probably a direct effect on the development of ischaemic damage.