Tretinoin has been shown to stimulate the synthesis of collagen in photo-aged human and hairless mouse skin. It has been suggested that this partial reversal of photodamage by tretinoin is a consequence of low-grade inflammation. The purpose of this study was to compare the effect of tretinoin with a number of irritants and peeling agents on collagen synthesis. Hairless mice were irradiated thrice weekly for 10 weeks with UVB. In the 10-week postirradiation period, the mice were treated topically five times per week with tretinoin (0.05%), glycolic acid (10%), benzalkonium chloride (1.0%), sodium lauryl sulfate (5%), croton oil (5%) and the water - propylene glycol vehicle. Microscopic measurements showed that the tretinoin-induced zone of new collagen was twice the depth of that induced by irritants or vehicle. The salt-soluble collagen content was determined by HPLC analysis of hydroxyproline levels. Type III procollagen was quantified by radioimmunoassay. Tretinoin-treated skin had increased amounts of collagen and type III procollagen whereas irritant- and peeling agent-treated skins were similar to vehicle-treated controls. Immunofluorescence studies were confirmatory. These results demonstrate that these agents, unlike tretinoin, do not have the capacity to enhance collagen synthesis. Therefore, it is likely that the effect of tretinoin does not depend upon irritation.