Recently, the Rhesus D-negative phenotype was identified in multivariate analysis as an important independent risk factor in gastric cancer patients who underwent curative resection. The distribution of Rhesus (Rh) D phenotype was investigated in a group of 268 patients who underwent gastrectomy for gastric cancer from June 1988 to April 1999. After excluding patients with short, potentially guaranteed follow-up or who deceased in the postoperative period, the prognostic significance of Rh D phenotype was evaluated in a subgroup of 239 patients by Cox regression model controlling for gender, age, site, histology, depth of tumor invasion (T), node metastasis (N), and type of resection (R). Two hundred and thirty-six patients (88.1%) presented an Rh+ phenotype and 32 (11.9%) presented an Rh- phenotype. A significant association was found between Rh D- phenotype and the presence of residual tumor after surgery (P = .01). The cumulative 5-year survival rate (95% confidence interval) was 37.3% (18.1-56.6) in patients with Rh- phenotype and 47.0% (39.2-54.4) in patients with Rh+ phenotype. The Rh D phenotype did not affect survival independently either in univariate analysis (P = .27) or in multivariate analysis (P = .55). The relative risk of death in D-negative versus D-positive patients decreased from 1.36 (95% confidence interval, 0.80-2.30) in univariate analysis to 0.84 (0.47-1.49) in multivariate analysis after controlling for depth of tumor invasion, lymph node metastases, and type of resection. Our study did not confirm the prognostic significance of Rh D phenotype in gastric cancer patients.