Purpose: Preoperative therapy for localized gastric cancer has considerable appeal. We hypothesized that, in a cooperative group setting, preoperative chemoradiotherapy would induce a 20% pathologic complete response (pathCR) rate. Combined-modality therapy quality, survival, and safety were secondary end points. Patients and Methods: Patients with localized gastric adenocarcinoma were eligible. A negative laparoscopic evaluation was required. Patients received two cycles of induction fluorouracil, leucovorin, and cisplatin followed by concurrent radiation and chemotherapy (infusional fluorouracil and weekly paclitaxel). Resection was attempted 5 to 6 weeks after chemoradiotherapy was completed. Quality of therapy was assessed with other end points. Results: Twenty institutions participated. Forty-nine patients were entered and 43 were assessable (12% stage IB; 37% stage II; and 52% stage III). The pathCR and R0 resection rates were 26% and 77%, respectively. At 1 year, more patients with pathCR (82%) are living than those with less than pathCR (69%). Grade 4 toxicity occurred in 21% of patients. Chemotherapy, radiotherapy, and surgery per protocol (including acceptable variations) occurred in 98%, 44%, and 63% of patients, respectively. A D2 dissection was performed in 50% of patients. Of 18 major radiotherapy variations, 17 were due to the lack of inclusion of the L3-4 vertebral interphase as prespecified. Conclusion: For localized gastric cancer, preoperative chemoradiotherapy strategy achieved a pathCR rate of more than 20% in a cooperative group setting. The quality of surgery improved (50% with D2 dissection) possibly because surgery was part of this trial. With some refinements, this preoperative chemoradiotherapy strategy is poised for a randomized comparison with postoperative adjuvant chemoradiotherapy in patients with gastric cancer.