To estimate the proportion of patients with gastric carcinoma that can be classified using the criteria of the fifth edition of the TNM system and to analyze which factors could be related to the finding of an adequate number of nodes. The influence of distinct factors that could influence the number of lymph nodes isolated was evaluated in 164 patients who underwent resection of gastric carcinoma. These factors included tumor size, surgical resection, grade, histological type, variability among the pathologists who analyzed the surgical specimens, and the surgeon's experience. The mean number of lymph nodes examined by the pathologists was 11.4 (10.12-12.66). Applying the criteria of the fifth edition of the TNM classification, only 31% of the patients could be correctly classified. A positive correlation was found between tumor size and the number of resected nodes (p = 0.0018). In addition, a greater number of lymph nodes were found in total gastrectomies than in subtotal gastrectomies (p = 0.034). No significant association was found with the pathologist who analyzed the surgical specimen or with the experience of the surgeon who performed the resection. The fifth edition of the TNM system is easily reproducible, although the number of lymph nodes required to evaluate metastatic node involvement is difficult to obtain in our environment. Our results suggest that a combined effort between surgeons and pathologists is needed to increase the number of patients that can be reliably staged with this TNM edition.