The cytokinesis blocked micronucleus assay is relatively insensitive to detect agents that predominantly induce excision repairable DNA lesions. However, it has been recently proposed that excision-repairable DNA lesions induced in G0/G1 phase can be converted to micronuclei by using inhibitors of the gap filling step of excision repair so that unfilled gaps are converted to double stranded breaks after S phase and micronuclei (MN) at completion of mitosis. As it has been recently demonstrated this process could be improved by combining cytosine arabinoside (ARA-C) and hydroxyurea (HU). In the present work, we have investigated the suitability of this new approach by studying its ability to detect excision repairable DNA lesions induced by 10 pesticides (alachlor, atrazine, cypermethrin, deltamethrin, fenpropathrin, fenvalerate, maleic hydrazide, paraquat, permethrin and trifluralin) and 3 well-known mutagenic agents (ethyl methane sulphonate, EMS; methylnitrosourea, MNU; and mytomicin C, MMC). Our results showed that the combination of ARA-C and HU substantially increased the level of MN in whole blood lymphocyte cultures, but it provided an excess of toxicity when further treatments, such as MNU, were performed. When ARA-C alone was used, the ARA/CBMN assay appeared to be highly sensitive and specific in detecting agents known to induce excision repairable DNA lesions. Thus, EMS and MNU but not MMC greatly induced DNA excision repair. On the other hand, alachlor, permethrin and, to a lesser extent, trifluralin and fenpropathrin also increased the ratio of excision repairable DNA lesions converted to MN. On the contrary, atrazine, cypermethrin, deltamethrin, fenvalerate, maleic hydrazide and paraquat did not induce excision repair.