Among workers employed at a nickel refinery in Norway between 1910 and 1977 an elevated risk of lung cancer has been demonstrated. A dose-related effect from nickel exposure has been identified, with the strongest gradient for water-soluble nickel. This pattern was recently confirmed in a nested case-control study with adjustment for smoking and potential occupational confounders. In the present study, updated cancer data were used to explore the risk by duration of work at the refinery and by exposure to different forms of nickel. Comparisons were made with the national male population (standardised incidence ratios) as well as internal reference groups (Poisson regression) under adjustment for age and smoking. The results confirmed earlier findings of a strong dose-related risk dependent on duration of work in production departments and cumulative exposure to nickel, most clearly seen for water-soluble nickel. Only slightly elevated risks were found among the unexposed and in the group with no experience from production or maintenance work. The risk associated with exposure to nickel chloride was similar to that for nickel sulfate. Analyses restricted to men exposed after 1967, with estimates based on personal monitoring of nickel in the breathing zone, showed the same risk pattern as for earlier years. Elevated lung cancer incidence was even suggested for workers with their first employment after 1978 when a lot of high exposure jobs were abandoned. The combined effect of exposure to nickel and smoking seemed to be in agreement with a multiplicative risk pattern.