The controversy on genotoxicity of molinate, an herbicide, has been reported in bacterial system, and in vitro and in vivo mammalian systems. To clarify the genotoxicity of molinate, we performed bacterial gene mutation test, in vitro chromosome aberration and mouse lymphoma $tk^{+/-}$ gene assay, and in vivo micronucleus assay using bone marrow cells and peripheral reticulocytes of mice. In bacterial gene mutation assay, no mutagenicity of molinate ($12-185{mu}g/plate$) was observed in Salmonella typhimurium TA 98, 100, 1535 and 1537 both in the absence and in the presence of S-9 metabolic activation system. The clastogenicity of molinate was observed in the presence ($102.1-408.2;{mu}g/mL$) of metabolic activation system in mammalian cell system using Chinese hamster lung fibroblast. However, no clastogenicity was observed in the absence ($13.6-54.3;{mu}g/mL$) of metabolic activation system. It is suggested that the genotoxicity of molinate was derived some metabolites by metabolic activation. Molinate was also subjected to mouse lymphoma L5178Y $tk^{+/-}$ cells using microtiter cloning technique. In the absence of S-9 mixture, mutation frequencies (MFs) were revealed $1.4-1.9{ imes}10^{-4}$ with no statistical significance. However, MFs in the presence of metabolic activation system revealed $3.2-3.4{ imes}10^{-4}$ with statistical significance (p<0.05). In vivo micronucleus (MN) assay using mouse bone marrow cells, molinate revealed genotoxic potential in the dose ranges of 100-398 mg/kg of molinate when administered orally. Molinate also subjected to acridine orange MN assay with mouse peripheral reticulocytes. The frequency of micronucleated reticulocytes (MNRETs) induced 48 hr after i.p. injection at a single dose of 91, 182 and 363 mg/kg of molinate was dose-dependently increased as $10.2{pm}4.7,;14.6{pm}3.9;and;28.6{pm}6.3;(mean{pm}SD;of;MNRETs/2,000;reticulocytes)$ with statistical significance (p<0.05), respectively. Consequently, genotoxic potential of molinate was observed in in vitro mammalian mutagenicity systems only in the presence of metabolic activation system and in vivo MN assay using both bone marrow cells and peripheral reticulocytes in the dose ranges used in this experiment. These results suggest that metabolic activation plays a critical role to express the genotoxicity of molinate in in vitro and in vivo mammalian system.