The primary cellular or molecular targets accounting for melanocytes loss in vitiligo are not clearly identified. To study a putative latent epidermal defect in the epidermis of vitiligo patients, we performed in vitro studies using cultured vitiligo melanocytes and keratinocytes transplanted on to a dead de-epidermized dermis according to a variant of Prunieras' technique. Control autologous constructs were made with keratinocytes and melanocytes of normal adult epidermis and vitiligo epidermis from perilesional skin. For heterologous reconstructs we combined vitiligo-derived melanocytes or keratinocytes with their normal phototype-matched counterpart. After 15 days of culture at the air-liquid interface, epidermal reconstructs were studied macroscopically and microscopically. Immunohistochemistry was performed using antibodies to TRP-1 and NKI-beteb. All heterologous and autologous reconstructs made with melanocytes and keratinocytes from vitiligo patients had a normal histology and ultrastructure. For vitiligo melanocytes or normal melanocytes submitted to the influence of vitiligo keratinocytes, immunophenotype and function (pigment production and transfer) were similar to normal controls. So, without additional noxious stimuli, we could not discriminate between melanocytes and keratinocytes as inducers of the disease. Our data suggest that the basic abnormality in vitiligo vulgaris needs extrinsic factors to be macroscopically revealed or requires a longer period of culture to develop. Our model will allow analysis of the various pathophysiological mechanisms of vitiligo, e.g. autoantibodies or oxidative stress, at the cellular, biochemical or molecular level