Plants have long been recognized for their pharmacological properties as evidenced by the extensive use of herbal medicines and tonics by many indigenous cultures. The earliest documentation of herb-based treatments appeared in Shen Nong Ben Cao Jing (The Divine Farmer’s Materia Medica) in 2737 BC. Despite a long history of herbal remedies, the scientific basis underpinning their efficacy in treating various maladies is lacking. Except for a limited number of successful cases, individual compounds isolated from their host species often lack expected therapeutic activities, a phenomenon previously attributed to matrix effect. In traditional Chinese herbal medicine, a prescription typically consists of a handful of ingredients mixed in a given ratio, whereas many of these ingredients are indeed referred to as efficacy-enhancing ingredients. Using a combination of quantitative metabolomics, mathematics, analytic chemistry, genomics and biochemistry, we decipher the molecular mechanisms underlying the matrix effect of traditional herbal remedies. We believe this project will potentially provide new systems-level insights into disease mechanisms, and further instruct new therapies of complex diseases through personalized cocktail medicine.