To develop human cell models for assessing the carcinogenic potential of chemicals, we established transgenic human cell lines and tested the sensitivity of known carcinogens using a cell transformation assay. A retroviral vector encoding an oncogenic allele of H-Ras (HBER) or c-Myc (HBEM) was introduced into human bronchial epithelial cells (HBE) immortalized by SV40 large T (LT) antigen, leading to increased cell proliferation but failing to confer a transformed phenotype characterized by anchorage-independent cell growth and tumor formation of immunodeficient mice. When these pre-transformed cells were treated with nickel sulfate (NiSO4), we found that it shortened the latency of malignant transformation at least by 19 wk in HBER cells or 16 wk in HBEM cells compared to vector control cells. Similarly, the latency of cell transformation was shorter by 15 wk in HBER cells or 9 wk in HBEM cells when cells were treated with benzo(a)pyrenediol epoxide (BPDE). HBER cells appeared to be more sensitive to TPA, NiSO4 or BPDE-induced cell transformation compared to human embryonic kidney cells expressing H-Ras (HEKR), implying that cell-type specificity is one of important factors determining the effectiveness of the assay. Using AFB1 and BaP as the representative pro-carcinogens, we also compared the efficiency of three different metabolic conditions in mediating cell transformation. Low dose chemical induction seems to be a prospective system used for metabolic activation of pro-carcinogens. Our findings provided direct evidence that a genetically modified human cell transformation model can be applied to the assessment of potent carcinogens.