This paper describes observed and expected mortality from cancers of the lung, larynx, nose, and kidney in a cohort of 54,509 nickel workers followed for 35 years. For analysis purposes the cohort was subdivided into men with and without service in one of the three high nickel dust areas of the operation: the Sinter Plants at Copper Cliff and Coniston, and the Leaching, Calcining and Sintering (LC&S) department at Port Colborne. At Copper Cliff Sinter Plant workers experienced three times the expected number of lung cancer deaths; the SMR rose steeply with increasing duration of service peaking at 943 with 10 to 15 years. A similar overall excess risk of lung cancer was seen in the smaller Coniston Sinter Plant again with an indication of an exposure risk gradient. Men in the LC&S department at Port Colborne also experienced a dose related excess risk of lung cancer death that rose to an SMR of 806 with 20 to 25 years of service. Nasal cancer deaths were increased at both the Copper Cliff Sinter Plant (6 deaths) and the LC&S department at Port Colborne (19 deaths), representing SMRs of 3,704 and 7,755, respectively, for this rare cancer. Laryngeal cancer and kidney cancer, both previously associated with nickel, were not in excess in these high risk groups. A further exploration of death from these causes in the lower exposure remainder of the cohort revealed an epidemiologically modest elevation in lung cancer death in miners (probably not nickel related) and parts of the Copper Refinery. No evidence of laryngeal cancer excess was found. Kidney cancer death was slightly increased in miners but as a whole was reasonably in line with expected. The results are consistent with either nickel subsulphide or nickel oxide, or both, being carcinogens.