The c-Myc oncoprotein is strongly implicated in B-cell neoplasms such as human Burkitt lymphomas and mouse plasmocytomas. Transgenic mice in which the myc gene is juxtaposed to an immunoglobulin enhancer (E(mu)-myc) also develop B-cell lymphomas, but relatively late in life. In addition, these neoplasms are invariably clonal, suggesting the involvement of additional mutations. Such mutations frequently affect the p53 tumour suppressor gene or its positive regulator Arf, hinting that inactivation of the p53 pathway might be the second hit required for the progression towards malignancy. However, even tumours arising in E(mu)-myc/Arf-null animals are thought to be clonal. This observation raised doubts whether overexpression of Myc in p53-null B-cell precursors is sufficient for tumorigenesis. To address this question, we have established a new, non-transgenic mouse model of B-lymphoma. This model is based on isolation of primary bone marrow (BM) cells, admixing them with packaging cells producing a Myc-encoding retrovirus (LMycSN), and subcutaneous injection into a host with which BM cells are syngeneic. Predictably, wild type BM cells infected in vivo by LMycSN were not tumorigenic. However, LMycSN-infected p53-null BM cells readily gave rise to B-cell lymphomas composed predominantly of late pro-B/small pre-B-cells. In these tumours, heavy chain gene rearrangements were analysed using two independent PCR-based assays. All neoplasms with DJ-rearrangements were found to be polyclonal. This result suggests that inactivation of p53 and overexpression of Myc is all that is necessary for the development of full-fledged B-lymphomas. Our model would also be instrumental in assessing the transforming potential of Myc mutants and in studying cooperation between Myc and other oncogenes.