A post-marketing study is an integral part of research that helps to ensure a favorable risk-benefit profile for approved drugs used in the market. Because most of post-marketing studies use observational designs, which are liable to confounding, estimation of the causal effect of a drug versus a comparative one is very challenging. This article focuses on methodological issues of importance in designing and analyzing studies to evaluate the safety of marketed drugs, especially marketed traditional Chinese medicine (TCM) products. Advantages and limitations of the current designs and analytic methods for postmarketing studies are discussed, and recommendations are given for improving the validity of postmarketing studies in TCM products.