Background EOX (epirubicin, oxaliplatin, and capecitabine) is one of the standard regimens for metastatic or locally advanced gastric cancer (GC). A new combination based on fractional docetaxel (low-TOX) has been developed in an attempt to increase the efficacy of EOX and reduce the heavy toxicity of classical docetaxel regimens. Methods Overall, 169 previously untreated GC patients were randomized between EOX ( arm A ) and low-TOX ( arm B ). The primary endpoint was progression-free survival (PFS), while secondary ones were overall survival (OS), overall response rate (ORR), disease control rate (DCR), and tolerability. The study was designed to detect a 35% (80% power at a two-sided 5% significance level) PFS increase with low-TOX and an interim analysis for futility was planned after the first 127 events. Results At the cut-off date of interim analysis, median PFS was 6.3 months 95% confidence interval (CI) 5.0–8.1 in arm A vs 6.3 months (95% CI 5.0–7.8) in arm B , without statistical difference. OS was comparable in the two arms: 12.4 in arm A (95% CI 9.1–19.2) vs 11.5 months in arm B (95% CI 8.6–15.0). ORR was 33% and 24%, while DCR was 68% and 67%, respectively. Treatment modification (91% vs 78%, P  = 0.017) and number of patients with CTC grade ≥ 3 adverse events (42 vs 35) were higher in arm B. Conclusions A triplet regimen based on the fractional dose of docetaxel achieves no improvement over EOX which remains a potential standard treatment in many patients with inoperable, locally advanced or metastatic GC.