Variability in the effects of the intraluminal suture method of middle cerebral artery occlusion (MCAO) in the rat has been a common and disadvantageous finding. Therefore, we systematically investigated the effects of suture type and rat strain on outcome. First, the clinical and neuropathological effects of permanent MCAO with either an uncoated or a silicone-coated nylon suture were studied over 7 days in Sprague-Dawley rats (n = 36 for each type of suture). Outcome was less severe with the uncoated compared with the silicone-coated suture (e.g. total cerebral infarct volume at 24 h before any fatalities was 119.9 +/- 79.8 mm3, cf. 183.0 +/- 36.5 mm3, n = 12 for each, P < 0.05; and overall mortality rate was 12.5% cf. 33%, respectively), but much more variable (coefficient of variation was 66.6% cf. 19.9%, respectively). Second, being more consistent in its effects, the silicone-coated suture was further studied in Wistar and Fischer-344 rats (n = 12 for each). Seventy-five per cent of the Wistar's died prematurely from gross hemispheric oedema. Motor deficit and extent of infarction in the Fischer-344 rats were both significantly greater compared with Sprague-Dawley rats (e.g. total cerebral infarct volume at 24 h in the former was 253.6 +/- 25.4 mm3, n = 11, P < 0.05), and more consistent (coefficient of variation was only 10.0%). It was concluded that the silicone-coated suture and the Fischer-344 rats strain produced the most consistent results and their novel combination provides a reliable acute stroke model.