The phytogenotoxicity of 2,4-dinitrotoluene (2,4-DNT) and 2,6-dinitrotoluene (2,6-DNT) was assessed using the Tradescantia micronucleus (Trad-MCN) bioassay. Tradescantia cuttings bearing young inflorescences were exposed for 6 h to 2,4- or 2,6-DNT amended water solutions up to their respective solubilities. The nominal concentrations were 0, 1.9, 3.8, 7.5, 15, 30, 60, 100, 150, 200 mg/l of 2,4-DNT, and 0, 7.5, 15, 30, 60, 90, 120, 180 mg/l of 2,6-DNT. Each treatment was repeated three or four times. Chemical concentrations in test solutions were analyzed prior to and after the exposure. Cadmium chloride (0-20 mM) was used as the positive control. Micronuclei (MCN) were scored in the tetrad-stage pollen mother cells. The MCN frequency (%), i.e. the number of micronuclei scored in 100 tetrads, was the measurement endpoint. Results indicated that both 2,4-DNT and 2,6-DNT were genotoxic with the minimum effective dose (MED) of 30 and 135 mg/l, respectively. Longer exposure (30 h) without recovery time at 150 mg/l of 2,4-DNT and 180 mg/l of 2,6-DNT did not induce significantly higher MCN frequencies.