Accurate staging of early breast cancer requires pathological assessment of axillary lymph node involvement. We evaluated the proportion of women receiving surgery for early-stage breast cancer who do not receive any lymph node biopsy (LNB) and factors associated with not receiving LNB. Patients receiving surgery for early-stage breast cancer (T1a/T1b/T1c/T2N0) during the period 2003-2005 were selected from the National Cancer Database. Patient sociodemographic, clinical, health insurance, and facility information was collected. Logistic regression was used to assess factors predictive of not receiving LNB. The number of women meeting study inclusion criteria was 184,050, 11% of whom did not receive any LNB. Compared with White patients, Black patients had greater likelihood [odds ratio (OR) 1.10, p<0.001] of receiving no LNB; there were no significant differences for Hispanic or other non-White patients. Individuals who were uninsured (OR 1.24, p<0.0005) or covered by Medicare at age <65 years (OR 1.29, p<0.0001) had greater likelihoods of no LNB compared with those with private insurance. Medicaid patients and Medicare patients >or= 65 years were not significantly different from private insurance patients. Compared with the youngest quartile of patients (age <or= 51 years), patients in the oldest quartile (age >or=73 years) were more than three times as likely (OR 3.30, p<0.0001) not to receive any LNB. We conclude that, while guidelines indicate that LNB may be considered optional in certain patient groups, it remains a key component in determining stage, and thereby prognosis and appropriate treatment options. These results indicate that significant disparities exist in sampling of axillary lymph nodes among women with early-stage breast cancer.