Of the 1,972 deaths investigated by autopsy at the Los Angeles County Hospital in 1951, 64 were attributed to bronchogenic carcinoma. However, in some of these cases the protocols did not contain sufficient data to substantiate the diagnosis or the tissue specimens were not conclusive evidence. Although the survey left no doubt that the lung was the most common site of primary carcinoma in the series studied, a wider application of the findings is limited by the facts that patients with bronchogenic carcinoma are more likely than others to die in hospitals and that selection of cases for autopsy depends on the interest of the physician and the consent of the patient's survivors.