To determine the incidence and factors associated with the development of arm edema in women who participated in the National Surgical Adjuvant Breast and Bowel Project (NSABP) study B-04.Between 1971 and 1974, the NSABP protocol B-04 randomized 1,665 eligible patients with resectable breast cancer to either (1) the Halstead-type radical mastectomy; (2) total mastectomy and radiotherapy to the chest wall, axilla, supraclavicular region, and internal mammary nodes if by clinical examination axillary nodes were involved by tumor; and (3) for patients with a clinically uninvolved axilla, a third arm, total mastectomy alone. Measurements of the ipsilateral and contralateral arm circumferences were to be performed every 3 months.There was at least one recorded measurement of arm circumferences for 1,457 patients (87.5% of eligible patients). There were 674 women (46.3%) who experienced arm edema at some point during the period of follow-up until February 1976. For radical mastectomy patients, total mastectomy and radiotherapy patients, and total mastectomy patients alone, arm edema was recorded at least once in 58.1%, 38.2%, and 39.1% of patients, respectively (p < .001) and at last recorded measurement in 30.7%, 14.8%, and 15.5%, respectively (p = <.001). Increasing body mass index (BMI) also showed a statistically significant correlation with arm edema at any time (p = .001) and at last assessment (p = .005).Patients who undergo mastectomy, including those whose treatment plans do not include axillary dissection or postoperative radiotherapy, suffer an appreciable incidence of arm edema.