Absorption of the herbicide paraquat (1,1'-dimethyl-4,4'-bipyridinium) by mixer-loaders and spray operators on a Sri Lankan tea plantation was assessed over five consecutive days of spraying. Beginning on the day before spraying started and continuing for each of the five spraying days and for seven days after the last day of spraying, 24-hour urine samples were collected from each of the workers. Potential dermal exposure was assessed during further applications of paraquat on the day after the last day of urine collection. For this purpose two spraying replicates were conducted that involved the handling or spraying of an amount of paraquat equivalent to the maximum used per day in the assessment of absorption. The mixer-loaders and spray operators incurred, on average, approximately equivalent amounts of potential dermal exposure (66 mg and 74 mg paraquat ion, resp.); however, the distribution of the exposure differed. About 86% of the total exposure experienced by the mixer-loaders was to the hands, whereas about 90% of the exposure of the spray operators involved their hands, legs, and feet, in approximately equal proportions. In both cases, 90% or more of the total potential exposure involved parts of the body that were normally uncovered. Despite the evidence of dermal exposure, no paraquat was detected in the workers' urine. This probably was due to the very low concentration of paraquat in the solutions used for spot spraying on tea plantations (0.3-0.4 g paraquat ion per litre), the high standard of personal hygiene exercised by the workers, and the low permeability of human skin to paraquat.