Given a plausible mechanism-based explanation of a pharmacological phenomenon, it is now feasible to build a virtual version of that explanation in software. The model mechanism and its generated phenomenon can be incrementally improved to become strongly analogous to actual counterparts. By so doing, we open a door to designing and conducting scientifically useful virtual experiments. The symposium will provide three examples of early stage virtual pharmacology, physiology, and therapeutics; show how the approach and methods can become scientifically and clinically productive; and describe how methods and tools are being adapted from engineering domains to improve transparency and credibility.