Herbal medicine, a multi-component treatment, has been extensively practiced for treating various symptoms and diseases. However, its molecular mechanism of action on the human body is unknown, which impedes the development and application of herbal medicine. To address this, recent studies are increasingly adopting systems pharmacology, which interprets pharmacological effects of drugs from consequences of the interaction networks that drugs might have. Most conventional network- based approaches collect associations of herb-compound, compound-target, and target-disease from individual databases, respectively, and construct an integrated network of herb-compound- target-disease to study the complex mechanisms underlying herbal treatment. More recently, rapid advances in highthroughput omics technology have led numerous studies to exploring gene expression profiles induced by herbal treatments to elicit information on direct associations between herbs and genes at the genome-wide scale. In this review, we summarize key databases and computational methods utilized in systems pharmacology for studying herbal medicine. We also highlight recent studies that identify modes of action or novel indications of herbal medicine by harnessing drug-induced transcriptome data. [BMB Reports 2022; 55(9): 417-428].