Heart-lung transplantation is a successful treatment for patients with cystic fibrosis and chronic respiratory failure. Patients are assessed for surgery when life expectancy is deemed short. This study assesses the ability of measurements of pulmonary function, blood gas levels, and nutritional status to predict survival of patients awaiting heart-lung transplantation and to assess the effect of heart-lung transplantation on survival. Sixty-seven patients with cystic fibrosis were accepted for heart-lung transplantation from 1985 through 1990. Each patient underwent tests of pulmonary function, exercise tolerance, blood-gas levels, and nutritional status. Cox regression was used to analyze the prognostic value of these data. Of the 67 patients accepted for heart-lung transplantation, organs became available for 30; 24 patients died waiting, and 13 patients were on the list December 31, 1990. Eight patients died after transplantation. Patients with above-average forced expiratory volume in 1 second (FEV1) (median, 17% predicted) at assessment were half as likely to die waiting (relative risk, 0.47; 95% confidence interval, 0.23 to 0.97) as patients with below-average FEV1. The relative risk of death for patients on the waiting list was 1.44 (95% confidence interval, 1.17 to 1.77) for each increment in PCO2 of 1 kPa. No effect of nutritional status on waiting list death was found, but seven of eight patients who died after heart-lung transplantation were below 80% predicted weight for height. Using a time-dependent analysis, heart-lung transplantation almost halved the risk of death (relative risk, 0.57; 95% confidence interval, 0.22 to 1.48). Postoperative actuarial survival was 79% (95% confidence interval, 63% to 95%).(ABSTRACT TRUNCATED AT 250 WORDS)