Docetaxel associated with oxaliplatin and 5-fluorouracil (FLOT) has been reported as the best perioperative treatment for gastric cancer. However, there is still some debate about the most appropriate number and timing of chemotherapy cycles. In this randomized multicenter phase II study, patients with resectable gastric cancer were staged through laparoscopy and peritoneal lavage cytology, and randomly assigned (1:1) to either four cycles of neoadjuvant chemotherapy (arm A) or two preoperative + two postoperative cycles of docetaxel, oxaliplatin, and capecitabine (DOC) chemotherapy (arm B). The primary endpoint was to assess the percentage of patients receiving all the planned preoperative or perioperative chemotherapeutic cycles. Ninety-one patients were enrolled between September 2010 and August 2016. The treatment was well tolerated in both arms. Thirty-three (71.7%) and 24 (53.3%) patients completed the planned cycles in arms A and B, respectively (p = 0.066), reporting an odds ratio for early interruption of treatment of 0.45 (95% confidence interval (CI): 0.18–1.07). Resection was curative in 39 (88.6%) arm A patients and 35 (83.3%) arm B patients. Five-year progression-free survival (PFS) was 51.2% (95% CI: 34.2–65.8) in arm A and 40.3% (95% CI: 28.9–55.2) in arm B (p = 0.300). Five-year survival was 58.5% (95% CI: 41.3–72.2) and 53.9% (95% CI: 35.5–69.3) (p = 0.883) in arms A and B, respectively. The planned treatment was more frequently completed and was more active, albeit not significantly, in the neoadjuvant arm than in the perioperative group.