Possible causes underlying the substantial differences in gastric cancer survival rates observed between Japan and the West were examined in a randomized trial comparing the Western R1 resection with limited lymphadenectomy and the Japanese R2 resection with extended lymphadenectomy. The effect of four factors associated with lymphadenectomy on microscopic tumor-node-metastasis (TNM) staging, and on stage-specific survival rates was assessed in 473 curatively resected patients. After application of extended lymphadenectomy, additional information on N status was available, only in R2 resections with up-staging to N2 status in 30% of patients. The calculated effect of this stage migration on known 5-year survival rates was as follows: an increase of 1% in TNM stage Ia, 2% in Ib, 7% in II, 15% in IIIa, and 15% in IIIb. A further increase in survival was observed by stage migration to N3 or N4 status, due to selective extension of lymphadenectomy to clinically overt metastases located outside the allocated level of clearance (contamination). Incomplete lymphadenectomy of N1- or N2-level stations indicated for dissection (noncompliance) demonstrates that more migration can occur when strictly adhering to the protocol. Examining more nodes per N level (diligence) induces even more migration, since the number of nodes that were histologically examined per N level correlated significantly with nodal status (lymph node-negative [N-] or lymph node-positive [N+]). These factors explain, at least partially, superior stage-specific survival rates after R2 compared with R1 resections, without a real survival benefit in individual patients.