To achieve sensitive and specific mechanism-based prediction of drug toxicity, the tools of systems pharmacology will be integrated using structured ontological approaches, analytics, mathematics, and statistics. Success of this effort is based on the assumption that a systems network that consists of drug-induced perturbations of physiological functions can be characterized. This network spans the hierarchy of biological organization, from gene to mRNA to protein to intracellular organelle to cell to organ to organism. It is populated with data from each of these levels of biological organization. These data, from disparate sources, include the published literature, drug development archives of all approved drugs and drug candidates that did not complete development, and various toxicity databases and adverse event reporting systems. The network contains interrelated genomics, transcriptomics, and metabolomics data, as well as organ and physiological functional data that are derived from the universe of information that describes and analyzes drug toxicity. Here we describe advances in bioinformatics, computer sciences, next-generation sequencing, and systems biology that create the opportunity for integrated systems pharmacology-based prediction of drug safety.