OBJECTIVES: The optimal number of lymph nodes that need to be analyzed to reliably assess nodal status in distal pancreatectomy for adenocarcinoma is still unknown. METHODS: Two hundred seventy-eight patients who underwent distal pancreatectomy for adenocarcinoma were retrieved from a retrospective French nationwide database. The relations between the number of analyzed lymph nodes and the nodal status of the tumor were studied. The beta-binomial law was used to estimate the probability of being truly node negative depending on the number of analyzed lymph nodes. Cox proportional hazard model was used for the survival analysis. RESULTS: The median number of analyzed lymph nodes was 15. There was a positive correlation between the number of positive lymph nodes and the number of lymph nodes analyzed. The curve reached a plateau at approximately 25 lymph nodes. The beta binomial model demonstrated that an analysis of 21 negative lymph nodes shows a probability to be truly N0 at 95%. N+ status was associated with survival, but the number of lymph node analyzed was not. CONCLUSION: At least 21 lymph nodes should be analyzed to ensure a reliable assessment of the nodal status, but this number may be hard to reach in distal pancreatectomy.