OBJECTIVE To estimate, using Medicare claims data, the outcomes in elderly Americans undergoing lung resection for lung cancer. METHODS We used discharge diagnosis and procedure codes in 1983 to 1985 Medicare hospital (part A) claims records to identify patients who underwent lung resection for lung cancer; we assessed perioperative, one-year, and two-year survival using Medicare enrollment file data. METHODS From a nationally random sample of 1,138,000 Medicare beneficiaries over 65 years of age, we identified 1,290 individuals who fulfilled our definition for lung resection for lung cancer. RESULTS Overall perioperative (30-day) mortality was 7.4 percent. Postoperative survival at one and two years was 69 percent and 54 percent, respectively. Male sex, older age, and pneumonectomy, as opposed to a lesser procedure, were associated with reduced perioperative and one-year and two-year survival. The adverse effect of advanced age on one-year and two-year survival following lung resection was not explained by the lower life expectancy of older individuals. CONCLUSIONS Medicare claims data can be used to estimate likely outcomes for elderly patients undergoing surgery for lung cancer. Expected outcomes vary with the patient's age, sex, and the type of surgical procedure performed.