The aim of this study was to compare the influence of irritant and allergic reactions on the barrier function of the skin and to correlate disturbances of barrier function with macroscopically visible changes.For this purpose, the skin of the volar side of the forearm was exposed to nickel sulphate (5%), sodium lauryl sulphate (5%), dimethyl sulphoxide (50%) and phenol (5%) during a period of 48 h. Distilled water served as control.The exposures were made in a test panel of 23 female subjects consisting of a group of 14 patients with positive skin reactions to nickel sulphate and nine controls in whom no evidence of skin disease was found.Both skin water vapour loss measurements and a 'visual scoring sytem' (based on erythema, fissuring and scaling) were used to assess the effects ofthe substances on the skin. The test sites exposed to nickel sulphate were also read for allergic reactions.In both groups all test substances caused an increase in skin water vapour loss in comparison with distilled water. The increase was found to be statistically significant for nickel sulphate, sodium lauryl sulphate and dimethyl sulphoxide (a = 005).Erythema was consistently elicited by the test substances, particularly by sodium lauryl sulphate. In addition, dimethyl sulphoxide regularly caused fine cracks (chapping).The correlation between erythema and the increase in water vapour loss after exposure ofthe skin to an irritant varied considerably from substance to substance and was poor for nickel sulphate (5%) and phenol (5%) and negative for dimethyl sulphoxide (50%). A significant correlation was found for sodium lauryl sulphate (5%), This implies that a linear relationship between barrier function impairment and damage to the viable cells of deeper layers ofthe skin may not be assumed beforehand,