Topical application of certain petroleum middle distillates (PMD) to mice produces skin tumors after long latency, and initiation/promotion protocols indicate that this effect is associated with their tumor promoting activity. Since induction of sustained, potentiated epidermal hyperplasia is predictive of promoting activity, five compositionally distinct PMD [hydrodesulfurized kerosene (API 81-07); hydrodesulfurized PMD (API 81-10); odorless light petroleum hydrocarbons; severely hydrotreated light vacuum distillate (LVD); and lightly refined paraffinic oil (LRPO)] were assessed for their effects on epidermal hyperplasia. PMD were administered (2 x/week for 2 weeks) to skin of CD-1 mice. Four quantitative biomarkers of epidermal hyperplasia were evaluated: epidermal thickness, number of nucleated epidermal cells per unit length of basement membrane, labeling (BrdUrd) index of epidermal cells, and induction of epidermal ornithine decarboxylase (ODC) activity. As positive controls, 12-O-tetradecanoylphorbol-13-acetate (TPA) and n-dodecane were utilized. PMD-induced skin irritation was evaluated visually and/or histopathologically. All five PMD produced dose-dependent, skin irritation and epidermal hyperplasia. On a weight basis the magnitude of the maximal PMD-induced effects was similar to that produced by n-dodecane, but > 1000-fold less than that produced by TPA. Epidermal hyperplasia and subacute skin irritancy produced by the five PMD were similar. Of the four short-term markers of tumor promotion assessed, labeling index and epidermal ODC activity were predictive of the relative promoting activities of those PMD for which tumorigenicity bioassay data are available, i.e., API 81-07 > API 81-10 > LRPO. An apparent discrepancy to the predictability of epidermal ODC activity occurred with LRPO:toluene [1:1 (v/v)]. This mixture is nontumorigenic, yet significantly induced epidermal ODC activity. This mixture, however, produced severe epidermal toxicity that precluded any meaningful analysis of short-term biomarkers in relationship to biological activity.