Gastric carcinoma is the second most common cause of digestive tumour-related death in Europe, North America and Asia. Today, the gold standard of treatment is still surgery, but outcomes to date are unsatisfactory. The Japanese Society for Research in Gastric Cancer (JSRGC) recommends the routine execution of splenectomy during gastrectomy. This recommendation is contested by western Authors because of increased morbidity and mortality without any real advantage in terms of survival. Patients treated for gastric cancer in our department between 1993 and 2002 were selected for this study. The 132 patients recruited were divided into two groups: a) those treated for gastric cancer without splenectomy; and b) those in whom splenectomy was performed in conjunction with gastrectomy. We analysed: the extent of lymph node dissection, the execution of the splenectomy, and the prognostic importance of factors relating to the patient, tumour and surgeon. Splenectomy was not associated with any increase in morbidity. Complications, especially of the septic type, and perioperative mortality were similar in both groups, and the same was true of survival at both 3 and 5 years. In our opinion, splenectomy should not be routinely combined with surgery for gastric cancer but could be considered for T3-T4 neoplasms or those localized in the upper two thirds of the stomach.