The relationship between self reported symptoms and subsequent all cause mortality was investigated in an occupational cohort consisting of 1224 nonsupervisory white male pulp and paper workers. Blood pressure, serum cholesterol, height, and weight were measured. Symptom reports were measured using the Cornell Medical Index (CMI) and the vital status of each participant was ascertained by reviewing company personnel records and death certificates. Results were compared to baseline characteristics available from a 1959 study of the cohort. Cohort mortality experience was compared to the mortality experience of the general United States population. Blood pressure, serum cholesterol, height, and weight were typical for worker age. Distribution of CMI scores showed a positive skew with 90 percent of participants having CMI scores less than 27. Application of two variable (age and CMI scores) and five variable (age, CMI, blood pressure, cholesterol, and Quetlets index) regression models based upon 24 years of followup yielded mortality rate ratio estimates of 1.23 and 1.24, respectively. Examination of expected deaths by category of predicted 24 year mortality risk based upon a discrete proportional hazards model of followup time, baseline age, and baseline CMI total score reasonably approximated observed data. Predicted risk values showed monotonic, but nonlinear effects of age and CMI scores upon risk. Among workers 40 years of age, successive 15 point CMI score increments yielded risk ratios of 1.210, 1.204, 1.198, and 1.191, while among workers 60 years of age, identical CMI score increments yielded risk ratios of 1.128, 1.110, 1.091, and 1.092. Within each age group the observed mortality rate among paper workers was higher, and the healthy worker effect was lower in the 50 percent of the cohort with high CMI scores. The authors conclude that there is a positive association between the number of reported symptoms at intake and subsequent mortality, independent of worker age and biologic risk at intake.