The authors review the available evidence on cancer risk associated with exposure to agricultural chemicals. Agricultural workers generally show a lower cancer mortality compared with other occupational categories. This observation is currently believed to be due to lower cigarette consumption. However, for some types of tumours (lymphoma, leukaemia, myeloma, soft tissue sarcoma, skin, prostate, brain and stomach tumours), mortality is higher among agricultural workers. The only chemical substances used in agriculture for which the IARC Monographs have established the existence of sufficient evidence of carcinogenicity for man are arsenical compounds and mineral oils; for other substances there is clear evidence of carcinogenicity in experimental animals, mostly in the absence of human data. In the case of exposure to phenoxyacetic herbicides, the available epidemiological evidence is contradictory, with excesses of non-Hodgkin lymphoma and soft tissue sarcoma reported in some studies but not in others. Cohort studies have been performed among insecticide production workers and spray operators (with excesses of lung tumour), and among grain processing workers (with excesses of non-Hodgkin lymphoma in particular). A number of case-control studies are also available, especially concerning tumours of the lymphatic and haemopoietic systems and ovarian tumours.