Topical application of phorbol ester tumour promoter evoked the stimulated generation of hydrogen peroxide in the mouse skin in vivo, probably resulting from its action on an associated non-target population of infiltrating inflammatory cells. This formation was demonstrated indirectly by hydrogen peroxide dependent 3-amino-1,2,4-triazole catalase complex formation and confirms the hypothesis and findings of Goldstein et al. (1983). In the assay model discussed, n-alkanes representing the range C6-C14 caused stimulated production of hydrogen peroxide but in mouse skin already containing inflammatory cells. It is proposed that the known tumour-promoting activity of dodecane and tetradecane is free radical mediated, and that this activity is expressed when these alkanes cause an inflammatory state such as occurs upon repeated dermal application. Free radical generation and possible tumour-promoting activity of lower alkanes is latent, due to lower irritancy and high volatilization, unless these alkanes are encountered in combination with an irritant. Skin explant studies demonstrated a priming potential of the phorbol ester tumour promoter to the induction of skin catalase by oxygen. This dose-dependent induction is also seen in untreated skin and may represent a possible compensatory response to oxygen toxicity. Such induction may confound the interpretation of other air-incubation studies in vitro. The generation of hydrogen peroxide was stimulated as a consequence of abrasion, a known tumour promoting agent, and this indicates that irritation, (and possibly the immune response) has a role in tumour promotion and in the variation in strain susceptibility to tumour promoters. Current techniques for the assay of enzymic activities specific for cytochrome P-450 and cytochrome P-448 failed to detect these in mouse skin. The presence of these activities in mouse hepatic tissue was demonstrated. Preliminary data is presented for a novel skin irritation screening test based on phorbol ester-stimulated oxygen consumption by inflammatory cells responding to the applied test material. The technique may offer a more precise and convenient measure of irritancy than that provided by the currently used Draize test.