The bioactivation of the pulmonary toxin 3-methylindole has been postulated to proceed via the formation of an imine methide. To test this hypothesis, the toxicity in mice of 3-methylindole has been compared to the toxicity of its perdeuteromethyl analog. Deuteration of the methyl group should slow the rate of production of the corresponding imine methide and diminish the toxicity of deutero-3-methylindole, if CH bond breakage occurs prior to or during the rate-determining step. In agreement with this hypothesis, deutero-3-methylindole was synthesized and was shown to be significantly less toxic (LD50 735 mg/kg) than 3-methylindole (LD50 578 mg/kg). Both compounds produced the same lesion at the LD50 dose, bronchiolar damage and mild alveolar edema, indicating that deuteration of 3-methylindole did not change the pathologic process. However, at a much lower dose (25 mg/kg), 3-methylindole produced a mild bronchiolar lesion whereas deutero-3-methylindole did not damage lung tissue. Additionally, administration of deutero-3-methylindole caused less pulmonary edema compared to 3-methylindole, as assessed by increased wet lung weights. Finally, the depletion of pulmonary glutathione by deutero-3-methylindole was considerably slower than depletion by 3-methylindole. The electrophilic imine methide has been postulated to be the intermediate which binds with and depletes glutathione. Therefore, the evidence presented here supports the involvement of an imine methide as the primary reactive intermediate in 3-methylindole-mediated pneumotoxicity.