OBJECTIVES: Administrative claims data are used to study BC treatment patterns and outcomes.  We critically review existing algorithms for identifying incident BC cases and disease stage at diagnosis within claims data.  METHODS: A literature search (1989-2012) using the terms: cancer [ti] AND (administrative OR claims) AND (assess [ti] or assessment [ti] or predict [ti] or prediction [ti] or identify[ti] or identification [ti] or validate[ti] or validation[ti] or agreement[ti] or agree[ti]) identified 76 articles with 27 testing algorithms. Of the 27, three provided algorithms for identifying incident BC cases and four algorithms classified BC disease stage at diagnosis. We examine the positive predictive value (PPV) of each algorithm and suggest revisions for improving the PPV. RESULTS: To identify incident BC, Warren used ICD-9 and treatment codes across inpatient and outpatient settings (PPV=80%), Freeman used a logistic regression with 36 diagnostic and procedural indicators (PPV=70%), and Nattinger used a 4-step process using ICD-9 and treatment codes (PPV=93%).   For disease stage, the Yuen and Cooper papers used diagnostic codes to distinguish regional from distant spread of cancer, with PPVs below 60%.  Smith used diagnoses, procedures, and medical visits to estimate equations for distinguishing stage IV from all other patients and stage III from patients with stage I/II.  Trade-offs between sensitivity and PPV are made using cut-points from the models to classify patients into disease stage. Nordstrom’s algorithm identified metastatic disease using three components: a diagnosis code for secondary neoplasm OR any metastatic chemotherapy agent OR no G-code for non-metastatic disease AND G-code for metastatic disease (PPV=81%). CONCLUSIONS: For identifying incident BC, we recommend updating the Nattinger algorithm with newer codes and fewer criteria for older patients likely to receive less aggressive care.  For disease stage, we recommend updating the Smith algorithm with additional codes from Nordstrom and testing higher cut-points to maximize PPV.