Advanced gastric cancer is an aggressive disease, with continued poor clinical outcomes. In the past two decades, many clinical trials have been conducted, testing chemotherapy agents in different combinations. However, many of these trials have employed aggressive regimens, which may be not suitable for some patients with advanced disease and poor performance status seen in clinical settings. In addition, geographic and ethnic differences in disease pathogenesis/biology, treatment strategies and drug metabolism make universal application of these trial results difficult. Here, we summarize the extant data on chemotherapy agents in gastric cancer, and try to deduce the best combination(s) that may be utilized in this disease, or that may be suitable for particular subgroup patient populations. Importantly, we discuss future directions - how we need to move away from testing various cytotoxic agent permutations and combinations, towards smarter trials that employ targeted therapies.