The scientific literature published over the period 1975-1991 on long-term health effects from prolonged exposure to pesticides has been reviewed, in order to document the state-of-knowledge on the mortality and morbidity of groups of the population exposed to pesticides. Specific aims of the review were to identify (1) which groups of the population have been more broadly surveyed, (2) which adverse effects have been consistently associated with specific pesticide exposures, (3) which are the pesticides of concern and (4) which health effects would require special attention in future research. The literature on acute effects and animal toxicology studies have not been considered at all in this monograph. In the period of interest, 440 papers have been published. Apart from 97 reviews and a small number of case reports, approximately half of the original investigations were of the case-control design (n = 108), while the remaining papers reported results from proportionate mortality (n = 10), cohort (n = 66) or cross-sectional studies (n = 51), carried out on pesticide applicators (n = 48), agricultural workers (n = 26) or people employed in the pesticide manufacturing industry (n = 50). Most of the case-control studies related to cases of cancer from various sites, especially myelolymphoproliferative disorders (MLP) and soft-tissue sarcomas (STS). When compared to the general population total mortality has been found to be consistently lower among pesticide manufacturers as well as among other groups of workers. This observation has been mostly attributed to the 'healthy worker effect' or, in the case of agricultural workers, to the healthier lifestyle of farm families. With the exception of deaths by accidental causes, non-cancer causes of death (mainly represented by cardiovascular diseases), were generally found to be less frequent than expected among manufacturers or users of pesticides, in particular among farmers. No consistent evidence of a global cancer mortality different from that of the general population has been reported among pesticide manufacturers or applicators. On the other hand, the papers examined have been strikingly consistent in reporting a low overall cancer risk among agricultural workers; life-style, clean air, low prevalence of smoking have been hypothesized so as to explain this observation. Numerous studies considered the possible link between exposure to phenoxyherbicides and occurrence of certain types of cancer, especially STS and MLP disorders.(ABSTRACT TRUNCATED AT 400 WORDS)