The appropriateness of substituting a sentinel lymph node biopsy (SLNB) and regional nodal irradiation (RNI) for an axillary lymph node dissection (ALND) in patients with residual lymph node (LN) disease following neoadjuvant chemotherapy (NAC) is unknown, and is currently under investigation in a national randomized clinical trial. Here, we compare overall survival (OS) between these two treatment approaches using the National Cancer Database (NCDB). We performed a retrospective cohort analysis of NCDB patients with non-metastatic, cT1-3, cN1 breast cancer, treated with NAC with residual disease in 1-3 LNs (ypN1), between 2006 and 2014. Patients were grouped into those treated with SLNB (defined as removal of ≤4 LNs) and RNI versus ALND (defined as removal of >4 LNs) and RNI. Patients were matched for all available patient, tumor, and treatment characteristics using propensity scores with inverse probability of treatment weighting (IPTW). OS was compared using Kaplan-Meier estimates. We identified 1377 eligible patients in the ALND group and 318 patients in the SLNB group, with median follow up of 44 and 36 months, respectively. Median number of removed LNs was 13 (IQR 9-17) and 3 (IQR 1-4) in the ALND and SLNB groups, respectively. Compared to the ALND group, patients in the SLNB group were more likely to be treated in the second half of the study period (85% vs 78%), undergo partial mastectomy (57% vs 49%), and have residual disease in a single LN only (69% vs 43%). IPTW-matching achieved adequate balance between the two groups with no significant differences in any of the matched variables. For the matched cohorts, SLNB was associated with significantly lower survival in both univariate and doubly robust multivariable analyses (MVA) (HR 1.7, 95% CI 1.3 – 2.2, P<0.001 for MVA), with estimated 5-year OS of 70% compared to 77% in the ALND group (P = 0.02). Exploratory subgroup analyses showed inferior outcomes for the SLNB group even in patients with a single positive LN. SLNB was comparable to ALND in patients with luminal A/B breast cancer with a single positive LN (HR 1.3, 95% CI 0.9 – 2.1, P = 0.12). To validate our methodology, we compared SLNB to ALND in ∼ 25,000 patients who met the eligibility criteria of ACOSOG Z0011 and found similar survival rates between these two cohorts (HR 1.01, 95% CI 0.9-1.1, P = 0.8). Our analysis suggests that SLNB alone, even when combined with RNI, may be inadequate treatment of the axilla in patients with residual nodal disease following NAC. While exploratory subgroup analyses suggest SLNB might be appropriate in patients with minimal residual nodal disease and favorable tumor biology, an ALND should not be routinely omitted in this patient population until its efficacy is confirmed by prospective trials.