This is a prospective study of 8006 men of Japanese ancestry, born from 1900 to 1919, who were questioned and examined from 1965 to 1968. A wealth of epidemiologic-based data was collected on these men, and they have been under continuous hospital surveillance for cancer since their examination. By September, 1988, which is the proposed renewal date for this 5- year proposal, it is estimated that there will be 1289 cancer incident cases diagnosed in the study population: 215 prostate, 202 lung, 190 stomach, 188 colon and 101 rectal cancers, among others. We plan to study the following with this cohort: 1) the association of serum uric acid, an antioxidant, with cancer by specific type; 2) a further look at the relation of serum cholesterol to colon cancer; 3) the association of urinary bladder cancer (based on 90 incident cases) with cigarette smoking, coffee, alcohol, and the intake of dietary vitamin A and C; 4) the relation of pulmonary function measurements to lung cancer risk; 5) the association of physical activity with cancer risk, especially colon cancer; 6) the identification of epidemiologic risk factors for occult prostate cancer found incidentally in deceased subjects in the cohort who have received protocol autopsies; and 7) the identification of epidemiologic risk factors for occult thyroid cancer found incidentally in deceased subjects. In these two autopsy-based studies, subjects with occult cancer will be compared with subjects without occult cancer who have received protocol autopsies done by the pathologist. In addition to these seven objectives, there are three others. From 1971 to 1975, 6860 subjects (85.7% of the original 8006 men) were re-examined and serum samples were collected. With this resource, nested case-control studies will be done using pre- diagnostic serum to determine the following: 8) the relation of serum carotene and tocopherol fractions to squamous cell cancer (which includes cancer of the mouth, hypopharynx, esophagus, larynx, and lung); and 9) the association of certain serum trace elements (iron, zinc, copper) with cancer by specific type. Finally, a 30% systematic sample of the original subjects was also chosen for re-examination from 1970 to 1972, at which time fasting plasma triglycerides and lipoprotein fractions were measured. The association of plasma triglycerides and high, low and very low density lipoprotein with cancer risk at specific sites will be studied in this sub-sample of 2095 men.