The Goseki grading system, based on intracellular mucin content and tubular differentiation, was originally created to study the influence of histologic type on the mode of extension of gastric carcinoma. The prognostic value of this grading system was subsequently proposed and even recently supported, but controversies still remain about this topic. We applied the Goseki system on 114 cases of node-negative primary gastric cancer and compared Goseki groups with the other clinicopathologic features of the patients. Statistical analysis showed a significant correlation between Goseki grading and Laurèn classification, but failed to reveal any prognostic significance for this grading system. We believe that Goseki classification should not be routinely used for prognostic purposes.