In the present study, a porcine model for controlled skin expansion was investigated to improve our understanding of epidermal and vascular responses following stretching. The model is of outstanding importance not only for the clinical use of tissue expansion but provides interesting data for skin physiology and oncology, too. Thirteen out of 15 animals, who underwent silicone tissue expander implantation showed good clinical results. In all of them, skin biopsies were taken at the end of a controlled tissue expansion procedure (final expander volumes 350 or 500 ccm): one tissue specimen was obtained from the centre of the expanded skin area and a second from the neighbouring but nonexpanded skin. The tissue specimens were immediately frozen in liquid nitrogen and processed to 4 microns thick acetone-fixed frozen sections. Lectin histochemistry and immunohistology were performed using the following techniques: direct and indirect immunofluorescence technique (DIFT, IIFT), immunoperoxidase technique (POX) with either 3,3'-diamino-benzidine (DAB) or 3-amino-9-ethyl-carbazole (AEC). The histochemical findings were supplemented by measurements of the number of vital epidermal cell layers, the epidermal thickness (microns), and the papillary vascular count per visual field. There was a significant diminuation of the vascular count (mean +/- S.D. = 55.0% +/- 12.5%; U-test: p less than 5%). By immunohistochemistry, a loss of the basal cell reactivity for the following antibodies was noted: ACAM (against calmodulin), K 8.12 (against keratins 13 +/- 16) and A51-B/H4 (against keratins 8, 14, 18). There was a remarkable increase of filaggrin expression in the uppermost spinal cell layers in expanded skin, which was most pronounced in those specimens with the shortest interval to the last fluid injection into the expander. We gained no evidence for alterations of the expression of suprabasal epidermal keratins, lectin binding sites (UEA I, PNA, ConA, WGA), and vascular lectin- and immunoreactivity due to tissue expansion. The subdermal capsule, which had formed around the silicone expander, was strongly vimentin-reactive. In conclusion, controlled tissue expansion is capable to change the basal cell phenotype--a feature which is shared with a number of conditions with increased proliferative activity and with the epidermis covering different skin tumours. The regular expression of suprabasal keratins and epidermal lectin binding sites provides evidence for a normal epidermal cell differentiation. Furthermore, the porcine skin is a reliable model for studying physiology and pathophysiology of human skin.