The relationship between serum ceruloplasmin level and cancer incidence was investigated in a case-control study nested within a longitudinal study of 39,268 Finns participating in the Social Insurance Institution's Mobile Clinic Health Examination Survey carried out in 1968-1972. During a median follow-up of 8 years, 766 cancer cases were identified. Ceruloplasmin levels were determined from stored serum samples collected at the baseline from these cancer cases and from two matched controls per case. The overall incidence of cancer was positively associated with serum ceruloplasmin level. The association was strongest for lung cancer and other cancers related to smoking and, consequently, in males. The smoking-adjusted relative risk of lung cancer among men was 4.3 (95% confidence interval (CI) = 1.8-10.6) in the highest quintile of serum ceruloplasmin as compared with that in the lowest quintile. The corresponding relative risks for cancers related to smoking combined, and for cancers not related to smoking were 3.9 (CI = 1.9-8.4) and 0.9 (CI = 0.6-1.5), respectively. The elevated risk of lung cancer at high concentrations of serum ceruloplasmin persisted after further adjustment for several potential confounding factors such as serum levels of vitamins A and E and selenium. The risk was stronger during the first 6 years of follow-up than later, and strongest during the first 2 years. The most likely explanation of the present results thus is that high serum ceruloplasmin levels in lung cancer are mainly due to occult cancer.