Classification changes for bronchioloalveolar carcinoma (BAC) by the WHO in May 1999 narrowed its definition. This study was undertaken in an attempt to characterize the impact of these changes on the epidemiology of BAC. This retrospective study involves data analysis from the population-based Cancer Surveillance Programs of three Southern California counties from 1995 to 2003. BAC cases diagnosed after May 1999 are compared with BAC cases before that time by clinicopathologic variables including survival. Incident cases (11,969) of non-small-cell lung cancer (NSCLC) were analyzed, including 626 cases of BAC (5.2%). Median overall survival (OS) for BAC patients diagnosed after May 1999 (> 53 months) was significantly improved over median OS for BAC patients before May 1999 (32 months; P = .012). This survival benefit remained after adjustment for sex, smoking status, and stage at presentation (hazard ratio for time of diagnosis before May 1999 compared with a diagnosis after May 1999 = 1.43; P = .015). Median OS for non-BAC NSCLC patients diagnosed before May 1999 (9 months) did not differ from the median OS of such patients afterwards (10 months; P = .09). This epidemiologic study is the first to demonstrate a survival advantage for BAC patients diagnosed after May 1999 compared with BAC patients diagnosed before this time-a finding that persists after adjustment for sex, smoking status, and stage at presentation. We believe that this observed survival benefit likely reflects changes in the revised 1999 WHO classification.