Eighty-ͣfive subjects possibly exposed to herbicide following aerial spraying of Tordon 101 and Weedone 2,4-DP and 159 subjects from a referent, unexposed community were interviewed regarding exposure and health symptoms. The a priori hypotheses that exposure would be associated with reported worsening of respiratory symptoms and not with a dummy symptom were supported by the data. The relative risk for respiratory symptoms was thirteen. An exploratory analysis of responses regarding 32 symptoms indicated a significant association of exposure with eight symptoms: cough, difficulty breathing, sinus congestion, runny nose, swollen glands, wheezing, dizziness, and peeling skin. Adjustment for age, race, sex, smoking status, and educational attainment did not alter these findings. Those exposed subjects reporting a worsening, within a month, of any of the eight symptoms significantly associated with exposure, constituted the "reactor" group. Reactors so defined reported greater duration of exposure than the non—reacting exposed subjects. Reactors tended to be more educated and better acquainted with the identity of the sprayed material than the non-reactor's. The extent of over-reporting bias was assessed using dummy symptoms. This study constitutes the first epidemiological investigation of acute effects of community exposure to these herbicide formulations and demonstrates the importance of this type of community surveillance.