The survival benefit of an extended versus standard lymphadenectomy for gastric cancer (GC) is often attributed to upstaging when more lymph nodes (LNs) are removed, i.e., stage migration. An extended lymphadenectomy is defined as 30 or more LNs examined, a surrogate for a D2 dissection. The aim of this study is to examine whether the survival benefit of extended lymphadenectomy persists when stage migration is not possible. The National Cancer Data Base was queried to identify patients with pathologic N3 (pN3, ≥7 positive LNs) gastric adenocarcinoma. Overall survival (OS) was compared by extent of lymphadenectomy (7-14, 15-29, and ≥30 LN) and stratified by T stage. Of 2101 pN3 patients, 419 (19.9%) had 7 to 14 LNs examined, 1164 (55.4%) had 15 to 29 LNs examined, and 518 (24.7%) had ≥30 LNs examined. Unadjusted three-year OS in the entire cohort was 24.6, 27.3, 30.5 per cent for 7 to 14 LNs, 15 to 29 LNs, and ≥30 LNs, respectively (P = 0.003). On adjusted survival analysis by stage for patients with pT1-T2N3 disease, removing ≥30 LNs significantly improved OS compared with removing 7 to 14 LNs (hazard ratio [HR] 2.45, 95% confidence interval = 1.25-4.82, P = 0.009). Extended lymphadenectomy may confer a survival benefit in select patients with pT1N3 and pT2N3 GC, highlighting the importance of the number of LNs examined rather than stage migration on survival. For the majority of the N3 population, pT3-pT4, the extent of lymphadenectomy did not significantly improve the OS.