BACKGROUND : Although the current nodal staging system for gallbladder cancer (GBC) was changed based on the number of positive lymph nodes (PLN), it needs to be evaluated in various situations. METHODS : We reviewed the clinical data for 398 patients with resected GBC and compared nodal staging systems based on the number of PLNs, the positive/retrieved LN ratio (LNR), and the log odds of positive LN (LODDS). Prognostic performance was evaluated using the C-index. RESULTS : Subgroups were formed on the basis of an restricted cubic spline plot as follows: PLN 3 (PLN = 0, 1-2, ≥ 3); PLN 4 (PLN = 0, 1-3, ≥ 4); LNR (LNR = 0, 0-0.269, ≥ 0.27); and LODDS (LODDS < - 0.8, - 0.8-0, ≥ 0). The oncological outcome differed significantly between subgroups in each system. In all patients with GBC, PLN 4 (C-index 0.730) and PLN 3 (C-index 0.734) were the best prognostic discriminators of survival and recurrence, respectively. However, for retrieved LN (RLN) ≥ 6, LODDS was the best discriminator for survival (C-index 0.852). CONCLUSIONS : The nodal staging system based on PLN was the optimal prognostic discriminator in patients with RLN < 6, whereas the LODDS system is adequate for RLN ≥ 6. The following nodal staging system considers applying different systems according to the RLN.