Randomized controlled trials (RCTs) have challenged the need for routine radiation therapy (RT) in women ≥ age 70 with favorable early stage breast cancer (BC) due to modest improvement in local control and no survival benefit. We studied practice patterns in RT among elderly women in the United States. We analyzed data from the National Cancer Database (NCDB) of women ≥age 70 diagnosed with T1 or T2 and N0 invasive BC treated with breast conservation surgery (BCS) between 2004 and 2014. Patients were divided into four groups: (1) no RT, (2) partial breast irradiation (PBI); (3) hypofractionation (HF); and (4) conventional whole breast RT (CWBI). Univariable and multivariable analyses (MVA) were performed to compare characteristics among the four RT groups. A subgroup analysis of women with favorable disease (T1N0 ER + HER2‐) was also performed with similar statistical comparisons. Of the 66,126 meeting eligibility, 9,570 (14.5%) had PBI, 16,340 (24.7%) had HF, and 40,117 (60.7%) had CWBI. Only 99 patients (0.15%) had RT omitted. Omission of RT increased marginally from 0.04% in 2004 to 0.24% in 2014. MVA identified older age (OR 1.18, CI 1.08‐1.28), more comorbidities (Charlson‐Deyo Score of 1) (OR 3.36, CI 1.29‐8.72), and no hormone therapy (OR 22.07, CI 5.79‐84.07) as more likely to have RT omitted. The use of HF increased from 3.9% to 47.0%, P < .001 with a concomitant decrease in CWBI from 88% to 41%, P < .001. MVA identified older age, treatment location, and omission of chemotherapy as associated with HF. No significant differences from the larger cohort were found among the T1N0 subgroup analysis. Despite RCT evidence, omission of RT was rare in the United States, suggesting that more effective outreach methods to disseminate clinical guideline information may be needed.