The ability of small molecules such as urushiol, present as a wax on the poison ivy leaf surface, to cause allergic contact dermatitis (rhus dermatitis) has fascinated immunologists for decades. Current dogma suggests that these epicutaneously applied catechol-containing molecules serve as haptens to conjugate with larger proteins via reactive o-quinone intermediates. These complexes are then recognized as foreign antigens by the immune system and elicit a hypersensitivity reaction. Phorbol ester can directly induce cultured keratinocyte (KC) intercellular adhesion molecule-1 (ICAM-1) expression via a protein kinase C (PK-C)-dependent mechanism. As urushiol is also a known PK-C agonist, we asked if topical application of a poison ivy/oak mixture could directly induce epidermal KC ICAM-1 expression. During the pre-erythematous phase of this reaction (4 to 20 hours), epidermal KCs expressed ICAM-1; this "initiation phase" preceded the appearance of activated memory T lymphocytes in the papillary dermis, and thus appeared to be nonlymphokine mediated. A near-contiguous cellular-adhesion molecular network was identified by ICAM-1 staining of basal KCs, dermal dendrocytes, and endothelial cells. During the second 24-hour period with the onset of erythema and edema, there was an "amplification phase" of more intense KC ICAM-1 expression coupled with relatively weak KC HLA-DR expression that coincided with dermal and epidermal T-cell infiltration. This suggests the presence of lymphokines, such as gamma interferon, during the amplification phase because of KC HLA-DR expression. On cultured KCs, urushiol directly induced ICAM-1 expression but not HLA-DR. Thus, in addition to functioning as an antigenic hapten, urushiol directly induces KC ICAM-1 expression. The KC ICAM-1 expression may then alter the dynamic trafficking of memory T cells in the epidermis, so as to initiate cutaneous inflammation in a nonantigen specific manner. This initiation phase is followed by T-cell infiltration and consequent lymphokine production that significantly amplifies the original stimulus. Thus much can still be learned about the molecular pathophysiology of this common type of cutaneous inflammation.