7559 Background: Concordant with surgical series, our prior analysis of US SEER data demonstrates improved survival for BAC patients with stage IIIB disease due to multiple lesions in the same lobe compared to other stage IIIB patients, and also for patients with ipsilateral multicentric BAC compared to those with distant metastasis. In order for the BAC staging system to accurately reflect survival for these distinct patient subsets, we propose a revised staging system for advanced BAC and validate this system using California Cancer Registry (CCR) data. Methods: A case-only analysis of incident BAC cases in CCR during 1999–2003 was conducted. Based on modeling from our prior SEER analysis, the T4 descriptor for multiple tumors in the same lobe was downstaged to T3, and the M1 descriptor for ipsilateral multicentric BAC was downstaged to T4. The primary outcome measured was overall survival (OS). OS univariate analyses were conducted using the Kaplan-Meier method; multivariate survival analyses were performed using Cox proportional hazards ratios. Results: 1909 incident cases of histologically-confirmed BAC were analyzed. Using the proposed criteria, 162 (25%) of the 654 advanced BAC cases were reclassified: 73 cases with multiple lesions in the same lobe as T3 (stage II T3N0M0 [n=53] or T3NXM0 [n=1], stage IIIA T3N1–2M0 [n=18], stage IIIB T3N3M0 [n=1]); 89 cases with ipsilateral intrapulmonary metastasis were reclassified as T4 (stage IIIB, T4N1- 3M0). Univariate OS analysis of this validation set reveals an improved fit for the proposed staging system compared to the existing staging system ( Table 1 ). Analysis by stage reveals improvement in this proposed prognostic model over the existing staging system on multivariate survival analysis after adjustment for relevant clinicopathologic variables ( Table ). Conclusions: Our proposed BAC staging system modifications more accurately reflect survival characteristics for this unique type of non-small-cell lung cancer. [Table: see text]