All-trans retinoic acid (tRA) was previously shown to be active in wrinkle reduction in the hairless mouse photoaging model. To address the questions of whether tRA also alters glycosaminoglycan (GAG) synthesis and whether observed wrinkle effacement can be attributed to changes in total GAG production, the effects of tRA on de novo GAG synthesis were examined in this model. Sulfated glycosaminoglycans (S-GAG) and the non-sulfated hyaluronic acid (HA) labeled with [3H]-glucosamine or [35S]-sulfate were found to diffuse differentially into the medium during the labeling period in the control animals (9% and 35% of total incorporated label for HA and S-GAG, respectively). Furthermore, the diffusion of HA into media was significantly changed after tRA treatment (from 9% to 24%), but no alteration was observed in the diffusion of S-GAG. Separation of epidermis and dermis indicated that the additional HA in medium after retinoid treatment primarily originated from the dermis. When incorporated label from the medium and skin fractions was combined, both labeling protocols revealed that 10 weeks of tRA treatment did not increase the total (medium plus skin) de novo synthesis of either HA or S-GAG. Wrinkle effacement as induced by retinoids in the photodamaged mouse skin therefore can not be related to an increased total GAG synthesis.