Résumé : A growing body of evidence from human and animal cancer cytogenetics studies indicates that aneuploidy is an important chromosome change in carcinogenesis. To understand the role of this genetic phenomenon during the first steps of an experimental cancer model, molecular and cellular techniques were combined. A sequential cytogenetic study of a modified Solt-Farber liver cancer model in the rat was performed to identify the importance of chromosome versus genome mutations. Male Wistar rats were initiated with diethylnitrosamine (DENA), followed by a 2-acetylaminofluorene exposure to select resistant hepatocytes. Chronic phenobarbital (PB) treatment was used to induce promotion. Cell proliferation was induced by a necrogenic dose of CCl4, administered during the selection period (Gerlans protocol) or 3 days before hepatocyte isolation (experimental protocol). In order to discriminate between genetic events causing chromosome breakage (clastogenic) and those that induce chromosome loss (aneugenic), isolated micronucleated hepatocytes (MNH) were analysed for the presence of a centromere in the micronucleus (MN). Non-radioactive in situ hybridization with a rat centromere satellite 1 DNA probe was applied. Our results show that the majority of the observed genetic changes, expressed as MN during different preneoplastic stages, were of clastogenic origin. However, the number of induced aneugenic hepatocytes increased markedly during the promotion period of the Gerlans protocol (~7-fold above control) and during PB exposure in the experimental protocol (~4-fold above control). Additionally, these stages were also characterized by an increased level of MN expression (20.3 < ? MNH < 32.8), in comparison with the initiation stage after DENA exposure (13.5 < ? MNH < 17.1). Although it is not yet clear if these genetic alterations have a causative nature in neoplastic liver transformation, the use of interphase cytogenetics certainly might lead to a better understanding of the genomic changes which occur during experimental hepatocarcinogenesis.