Lung and nasal sinus cancer death rates among employees of a nickel refinery in South Wales were examined from the standpoint of quantifying the relationship between exposure and the risk of these two cancers. Areas in the refinery associated with high risk were identified by using a matched case-control approach, and a simple index of exposure for each man was constructed accordingly based on the duration of time spent in these areas. The dependence of relative risk and absolute excess risk for lung and nasal sinus cancers on age at first employment, calendar period of first employment, time since first employment, and duration of exposure in high-risk areas was then analyzed. The relative risk for nasal sinus cancer increased sharply with increasing age at first exposure but remained roughly constant throughout the period of follow-up, while that for lung cancer was independent of age at first exposure and dropped sharply with increasing time since first employment. The implications and limitations of these analyses are discussed in relation to the multistage theory of carcinogenesis and occupational risk assessment.