It has been suggested that the chick chorioallantoic membrane (CAM) could be used as an alternative to the rabbit eye test for irritation. Various authors have examined different methods for investigating the irritant potential of materials to the CAM. We used the CAM method as described by Luepke, to assess the effect of 34 materials of known in vivo eye irritancy, including fully formulated products ranging from skin creams to industrial detergents. Test materials were applied undiluted to the CAM on day 10 of incubation with rinsing after 20 seconds of contact. Effects to the CAM including vasodilation and haemorrhage were assessed for up to 5 min after application. A comparison was made between the rank order of effect on the CAM and the known ranking for in vivo eye irritancy. There was in general only a limited correlation between in vitro and in vivo results but a reasonable distinction between irritant materials and those of only very slight irritancy was observed. Among products of a similar type there was a better correlation between results from the CAM and in vivo eye results than when all materials were considered together. These data indicate that this method would only have a limited application as a screening procedure for eye irritation potential or as part of a battery of in vitro tests but so far has not been found to be superior to cell culture tests in our laboratory.