While carcinoma of the stomach is decreasing in incidence in the United States, it is still a major cause of cancer death. But gastric neoplasms are not decreasing in some other geographic areas. According to some studies, 30% of all cancer in the U.S.S.R. originates in the stomach. The rate of gastric neoplasms is greatest in Japan, and over 54% of all cancer in the male population arises in the stomach. The peak age for development of stomach cancer is between 70 and 80 years; over 60% of all stomach cancer is diagnosed in patients between the ages of 60 and 70, while more than 10% is found in those over 80. The main hope for cure at this time rests with surgical treatment. However, despite increased use of surgery, the 5-year survival rate of approximately 13% for patients diagnosed during 1955-59 has not improved to any degree since that time. The major drugs commonly used to treat gastric cancer are 5-fluorouracil (5-FU) and mitomycin C. Controversy still exists concerning the optimum method for administering 5-FU, the most frequently used drug in the United States. The standard loading-course method was attended by a high risk of severe toxicity and drug-related deaths. Several variations of the loading course have evolved. Currently, the Mayo Clinic group uses a 5-day course of 13.5 mg 5-FU/kg repeated every 5 weeks, wHh therapy interrupted if stomatitis or diarrhea develops; with this regimen the drug-related mortality rate was reported to be less than 1%. Studies have shown that 5-FU plus radiotherapy can enhance survival in patients with locally unresectable disease. The overall objective with 5-FU is 20–25% with an average of 4–5 months' duration of response. Despite the many patients treated with 5-FU, rarely has a systematic analysis been done of factors such as age, sex, disease-free interval, histologic grade of the tumor, or sites of metastases, which might predispose to a favorable or unfavorable response. In Japan the most commonly used drug for treatment of gastric cancer is mitomycin C, the second most frequently used drug in the United States. The overall objective response rate with mitomycin C is between 20 and 30%, with the higher response rates being reported in the Japanese data. The average duration of response ranges from 1 to 3 months. The nitrosoureas [1,3-bls(2-chloroethyl)-1-nitrosourea (BCNU), 1,3-cis(2-chloroethyl)-1-nitrosourea (CCNU), and methyl CCNU (MeCCNU)] have shown some evidence of activity against gastric cancer. BCNU has yielded an objective response rate of 18% (6/33) and an average duration of response of 4.5 months in gastric cancer patients, most of whom had no prior therapy. Adriamycin recently has been shown to have some antitumor activity, with an approximate response rate of 25%. Combination approaches have been more successful in stomach cancer than in any other gastrointestinal neoplasm. The Japanese have reported higher response rates with a combination of 5-FU, mitomycin C, and cytosine arabinoside. The therapeutic activity of this combination has been confirmed at the Memorial Sloan-Kettering Cancer Center. The Mayo Clinic has observed a 41% response rate with the combination of 5-FU and BCNU, with survival superior to a concomitant control of both single agents. More recently, the combination of 5-FU and MeCCNU also showed these high response rates. The utilization of chemotherapy as an adjuvant to surgery has been extensively evaluated in Japan, the U.S.S.R., and the United States. The Japanese have studied predominantly mitomycin C alone or in combination. While a few results are positive, most studies have not shown a meaningful survival gain. In the United States, studies with triethylenethiophosphoramide and 5-fluoro-2′-deoxyuridine have all been negative, but all utilized short-term regimens of chemotherapy; current trials involved long-term administration of MeCCNU plus 5-FU. The past experience with surgical adjuvant chemotherapy neither confirms nor denies the utility of this approach. However, this lack of a definitive answer, which is largely due to faulty study designs and possibly to inadequate chemotherapy, does permit some insight into certain principles that may be critical in designing future combined modality therapies.