Detergents are well known irritating agents in human as well as in animal,models. Using a murine keratinocyte cell line (HEL30) changes in the interleukin-1 alpha profile were characterized in response to three non-ionic detergents, all widely used in the cosmetics industry. The compounds used in this study were the most active (dodoxynol-9, Delta-9), moderate (polyglyceryl-4-lauryl ether, PEL) and mild (PEG-20-glyceryl ricinoleate + ricinoleamide DEA, PEG) in inducing cytotoxicity, measured as lactate dehydrogenase leakage and de novo protein synthesis, on the same cell line after 2 hr of treatment. All of the surfactants tested were able to induce IL-1 alpha production both at a secretory and cell-associated level. However, in order to achieve a similar IL-1 production different concentrations of surfactants were necessary. It was possible to calculate an EC(50) for IL-1 alpha release of 52.9 mu g/ml for Delta-9, of 293.7 mu g/ml for PEL and of greater than 9000 mu g/ml for PEG, At the concentration of 30 mu g/ml no release could be detected even after 24 hr of treatment with PEL or PEG. A time-course experiment also showed significant amounts of IL-1 alpha 20 min after treatment with Delta-9. These data confirmed Delta-9 as the most potent of the three non-ionic detergents tested in inducing IL-1 alpha release. The surfactants were also tested in vivo using the modified Draize test. Once again Delta-9 was the most active, followed by PEL and PEG. Considering the key role of IL-1 in the inflammatory response, the release of this cytokine by keratinocytes in vitro could be used as a more specific (in comparison with classical cytotoxic markers) and early marker to determine the irritant potential of water-soluble chemicals.