ALTHOUGH pharmacological doses of retinoic acid (RA) have a wide variety of actions in vivo(1), experimental difficulties have prevented a definitive assignment of its physiological functions, We recently made a dominant-negative retinoic acid receptor (RAR) by a single amino-acid substitution(2) which creates a dominant-negative thyroid hormone receptor(3). The mutated RAR efficiently inhibited the endogenous activities of RARs (alpha, beta, gamma)(2). Thus, targeted expression of the mutated receptor should reveal RA functions during organogenesis by blocking RA signalling in the tissues concerned. To address this possibility, we expressed the dominant-negative RAR in the epidermis, a potential target organ of RA(4). We report here that the resultant transgenic mice exhibited dramatic suppression of epidermal maturation, demonstrating the requirement of RA in normal skin development.