Purpose: To assess the maximum tolerability of a combination of S-1 and preoperative radiotherapy and to evaluate the feasibility and activity in patients with locally advanced rectal cancer. Methods: Patients (n = 30) with adenocarcinoma of the middle or lower rectum were enrolled in a phase I (n = 9) and/or phase II (n = 21) trial. A total dose of 45 Gy was delivered in 25 fractions over 5 weeks, and S-1 was orally administered twice a day on days 1–14 and 22–35. Surgical resection was scheduled 4–8 weeks after the completion of chemoradiation. Results: In phase I, the recommended dose (RD) of S-1 was 80 mg/m2/day, and the maximum-tolerated dose was never reached. A total of 27 cases, including the 6 RD cases in phase I, were enrolled in phase II. In phase II, a pathological complete response (pCR) was observed in 6/27 patients (22%), pathological downstaging was observed in 21/27 patients (78%), and a tumor volume reduction of 69 ± 22% was obtained. These results were similar to the previously reported pCR rates of 16–18%, pathological downstaging rates of 49–59%, and tumor volume reduction of 68% after chemoradiotherapy with capecitabine. Grade 3 adverse events consisted of one case of leukopenia (4%), 2 cases of anemia (7%) and 3 cases of diarrhea (11%). Overall, the adverse events were very mild. Hand-foot syndrome was not observed. Conclusion: The efficacy of chemoradiotherapy with S-1 seems to be equivalent to the efficacy reported for chemoradiotherapy with capecitabine, but the adverse events were much milder, although further study is warranted.