A cohort study that comprised 533 men was carried out to evaluate the risk of cancer incidence among workers exposed to chloroprene between 1966 and 1997 in a production plant in Isere (France). The risk of cancer was evaluated for the period 1979-1997 from data recorded by the cancer registry in the department of Isere. Standardised incidence ratios (SIR) were calculated using the whole population of the department as reference. An excess of risk of all malignant neoplasms in the cohort was shown (SIR, 1.26; Cl, 0.88-1.77). This excess concerned cancers of the lung (SIR, 1.84; CI, 0.84-3.49) and of the head and neck (SIR, 1.89, CI, 0.87-3.59). Where lung cancer was concerned, there was an increase in risk depending on the length of exposure. This relationship was inverse for the head and neck. The excess of risk of lung cancer was the highest in people not exposed or only slightly exposed. For head and neck cancer, this excess was the highest in people with medium exposure. Our results did not confirm the excess of risk of liver cancer discovered during other cohort studies aiming to evaluate the role of chloroprene. The variations in the excess of the risk of cancer of the lung, head and neck were hardly compatible with a professional risk factor. (C) 2001 Elsevier Science Ireland Ltd. All rights reserved.