At the Mayo Clinic, from 1970 through 1979, 84 patients (52 males and 32 females) had abdominal exploration for primary gastric lymphoma. All patients were observed a minimum of 5 years or until death. The histologic findings for all 84 patients were reviewed. Forty-four patients had "curative resection," and 40 patients had either biopsy alone or a palliative procedure. The probability of surviving 5 years was 75% for patients after potentially curative resection and 32% for patients after biopsy and palliation (p less than 0.001). The operative mortality rate was 5% overall and 2% after potentially curative resection. Increased tumor size (p less than 0.02), increased tumor penetration (p less than 0.01), and lymph node involvement (p less than 0.02) decreased the probability of survival, whereas histologic classification did not affect survival. Radiation therapy after surgery did not significantly affect the survival rate for the entire group or the survival rate for patients who had potentially curative resection. Resectability was associated with increased patient survival--independent of other prognostic factors--when our experience was analyzed by the Cox proportional-hazards model (p less than 0.005). It was concluded that an aggressive surgical attitude in the treatment of primary gastric lymphoma is warranted. The role of radiotherapy remains in question.