Guanazole and aphidicolin were chosen as candidates in the search for a selective, non-genotoxic inhibitor of DNA replication which could be used instead of hydroxyurea to measure DNA repair synthesis in rat hepatocyte primary cultures by liquid scintillation counting. The genotoxicity of these 3 chemicas was studied using the Salmonella/liver homogenate assay and the autoradiographic UDS test in hepatocytes. Hydroxyurea was positive in both of these assays. Guanazole and aphidicolin did not induce DNA repair in hepatocytes. Aphidicolin was not mutagenic for Salmonella typhimurium, whereas guanazole increased the revertant numbers of strain TA102 slightly. The incorporation of [3H]thymidine was measured by liquid scintillation to determine DNA repair induced by 2-acetylaminofluorence (2-AAF), aflatoxin B1, benzo[a]pyrene, cyclophosphamide, H2O2, 6-hydroxydopamine, N-methyl-Nâ²-nitro-N-nitrosuguanidine (MNNG), methylnitrosourea (MNU), 4-nitroquinoline-N-oxide and UV irradiation in the presence of either 10 mM hydroxyurea, 15 mM guanazole or 0.015 mM aphidicolin. Aphidicolin had an inhibitory effect on DNA repair. Except for the 3 chemicals mentioned below, the sensitivity of the DNA repair measurement was the same, no matter whether hydroxyurea or guanazole was used to inhibit replicative DNA synthesis. In the presence of hydroxyurea, DNA repair synthesis was found at lower concentrations in the case of aflatoxin B1, due to differences in the solvent control values, and in the case of H2O2, possibly due to a synergistic effect between hydroxyurea and H2O2. Guanazole allowed the detection of DNA repair induced by MNNG at lower concentrations, probably because of an antagonistic effect between hydroxyurea and MNNG. Based on these results, it was concluded that guanazole, but not aphidicolin, could be used instead of hydroxyurea to measure DNA repair sythesis by liquid scintillation in rat hepatocyte primary cultures. Although guanazole does not completely fulfill the criteria for an ideal DNA replication inhibitor, it has the advantage of being less genotoxic than hydroxyurea, and also appears to have a smaller potential to falsify the results by interacting with the test compounds. © 1990.