Abstract: Mastectomy is used to treat one third of the nearly 180,000 women diagnosed with breast cancer in the United States annually. In this study, we use population-level data from multiple years of the Surveillance, Epidemiology, End Results (SEER) database to further define patient, tumor, and geographic characteristics associated with immediate and early-delayed breast reconstruction. Population level de-identified data for the years 1998 to 2002 were extracted from the National Cancer Institute’s (NCI) SEER cancer database. All female patients who were treated with mastectomy for a diagnosis of ductal and/or lobular breast cancer (including Paget disease) were included. The primary end point of interest was odds of reconstruction. Multivariate analysis was performed to control for patient demographic and oncologic characteristics. A total of 52,249 patients met the inclusion criteria. Reconstruction was performed in 8,446 patients (16.2%). Odds of reconstruction varied by region from 0.60 (Seattle) to 2.81 (Atlanta). African Americans were noted to have a significantly lower likelihood of reconstruction when compared with Caucasian patients (OR 0.60 versus 1.00). Patients living in nonmetropolitan regions were also significantly less likely to undergo reconstruction. Receipt of radiation therapy was also negatively correlated with likelihood of reconstruction. In this multicenter, multiyear analysis of factors associated with immediate or early-delayed reconstruction after mastectomy, we demonstrate that younger age, white race, metropolitan locale, and lower stage disease were all independently associated with higher likelihood of reconstruction. This information provides insight into breast reconstruction utilization and will help guide future studies to understand how these factors affect patient and physician decision-making.