In this study we define the proliferative compartments of in vivo human epidermis, using specific antibodies related to cell differentiation (beta 1 and beta 4 integrins and K1/K10 differentiation keratins) and cell cycle (proliferating cell nuclear antigen [PCNA]) in combination with flow cytometric quantitation of the DNA content and optical characteristics of the cells. The beta 1 integrin (CD29) marked both of the potentially proliferative subsets in normal epidermis. One subset of normal epidermis is CD29+ K1/K10-, which was predominantly basal, and found to be comprised of slow cycling, small cells with primitive cytoplasmic organization. The vast majority (95.5%) of these cells were in a quiescent state (G0/early G1) as indicated by their lack of the cyclin, PCNA. The other proliferative subset of normal epidermis was CD29+ K1/K10+, which was suprabasal and occasional basal, highly proliferative, larger in size, and which exhibited a more complex cytoplasmic structure. Because early differentiation (K1/K10 expression) has begun in the CD29+ K1/K10+ subset, it is highly likely that they represent the proliferative population which is capable of transiently amplifying itself before terminal differentiation. Within lesional psoriatic epidermis, similar proliferative cell populations were present as in normal epidermis, and the hyperproliferative defect was localized to the beta 1 and beta 4 integrin+, K1/K10- populations, which in normal epidermis is basally located and quiescent with regard to cell cycle. In psoriatic epidermis, a six- to sevenfold increase in the number of cells in the S/G2+M phase of cell cycle was found among CD29+ K1/K10- cells (p < 0.05). Furthermore, all lesional K1/K10- cells showed high PCNA positivity, indicating that all these cells had been recently induced into cell cycle. By contrast, the proportion of cycling cells among lesional psoriatic CD29+ K1/K10+ keratinocytes was similar to normals. Anti-HLA-DR, CD45, and vimentin antibodies were used to concomitantly track the proliferative states of Langerhans cell, melanocyte, and infiltrating leukocyte populations. In normal epidermis, the cycling fractions (cells in S/G2/M phase) of these cells were similar to the CD29+K1/K10- keratinocytes, whereas in lesional epidermis their cycling pools were increased relative to normal, but not so much as the proliferative fractions of psoriatic CD29+ K1/K10- keratinocytes. These data demonstrate the use of simultaneous analysis of integrin expression, differentiation keratins, cyclin, cell cycle status, and optical characteristics of freshly isolated human epidermal cells. Such analysis allowed the physical identification and quantification of cy cling populations in normal human skin, and has enabled the precise location of the primary epidermal proliferative defect in psoriasis.