Purpose To use patient self-report to provide more valid estimates of whether radiotherapy (RT) is underutilized than possible with registry data, as well as to evaluate for disparities and the influence of preferences and provider interactions. Methods We considered 2,260 survey respondents who had nonmetastatic breast cancer, were age 20 to 79 years, were diagnosed between July 2005 and February 2007 in Detroit and Los Angeles, and reported to Surveillance, Epidemiology and End Results (SEER) registries (72% response rate). Survey responses were merged with SEER data. We assessed rates and correlates of RT receipt among all patients with invasive cancer receiving breast-conserving surgery (BCS) and among patients undergoing mastectomy with indications for RT (ie, positive lymph nodes or T3-4 tumors). Results Among 904 patients undergoing BCS with strong indications for RT, 95.4% received RT, and 77.6% received RT among the 135 patients undergoing mastectomy with strong indications (P < .001). Among 114 patients undergoing BCS with weaker indications (ie, elderly) for RT, 80.0% received treatment, and 47.5% received RT among the 164 patients undergoing mastectomy with weaker indications (T1N1, T2N1, or T3N0 disease; P < .001). On multivariate analysis, surgery type (P < .001), indication strength (P < .001), age (P = .005), comorbidity (P < .001), income (P = .03), patient desire to avoid RT (P < .001), level of surgeon involvement in decision to have radiation (P < .001), and SEER site (P < .001) were significantly associated with likelihood of RT receipt. Conclusion RT receipt was consistently high across sociodemographic subgroups after BCS but was lower after mastectomy, even among patients with strong indications for treatment, in whom clinical benefit is similar. Surgeon involvement had a strong influence on RT receipt.