The reviewed data document expression of both COX-1 and COX-2 in human and mammalian skin as well as up-regulation of COX-2 in specific types of epidermally- derived precancerous and cancerous lesions. In a physiologic context, some evidence suggests that COX-2 expression may be involved in homeostatic regulation of normal keratinocyte differentiation. In a pathophysiologic context, increased COX-2 expression in keratinocytes occurs following exposure to UV light and chronic exposure to UV light is associated with the generation of both actinic keratoses, and squamous cell carcinomas (reviewed in [47]). It is worth noting that both actinic keratoses and squamous cell carcinomas also show mutations in the p53 tumor suppressor gene [48]. Consequently, at least several gene alteration, on of which involves the COX-2 gene, appear to be required for the phenotypic epidermal squamous carcinoma cell. The fact that UV -induced basal cell carcinomas do not appear to express COX-2 to the same extent as the squamous cell carcinomas suggests that COX-2 expression or up-regulation is not necessarily a required property of the carcinoma phenotype. AT the same time, the more aggressive behavior of squamous cell carcinomas compared to basal cell carcinomas may have something to do with COX-2 expression since a relationship between COX-2 expression and metastatic potential has been demonstrated in human colon cancer cells [49]. In a therapeutic context, the data suggest that COX-2 inhibitors may have a role in suppressing the appearance of epidermal neoplasms (e.g. actinic keratoses, squamous cell carcinomas, and keratoacanthomas) that demonstrate increased COX-2 expression. This has already been shown in a murine model. However, the observed suppression of carcinogenesis has never been complete, and this finding suggests, as already mentioned, that other factors may be more seminal to the ultimate appearance of the different types of skin cancers. Nevertheless, other control points in the generation and action of PGE2 or other cyclooxygenase-derived eicosanoids may merit attempts to develop other types of chemotherapy or chemoprevention; these would include inhibitors of cytosolic phospholipase A2 as well as EP receptor antagonists. Future studies will no doubt document the validity, or lack thereof, of these investigative approaches.