Identification and approval of new drugs for use in patients requires extensive preclinical studies and clinical trials. Preclinical studies rely on in vitro experiments and animal models of human diseases. The transferability of drug toxicity and efficacy estimates to humans from animal models is being called into question. Subsequent clinical studies often reveal lower than expected efficacy and higher drug toxicity in humans than that seen in animal models. Microphysiological systems (MPS), sometimes called organ or human-on-chip models, present a potential alternative to animal-based models used for drug toxicity screening. This review discusses multi-organ MPS that can be used to model diseases and test the efficacy and safety of drug candidates. The translation of an in vivo environment to an in vitro system requires physiologically relevant organ scaling, vascular dimensions, and appropriate flow rates. Even small changes in those parameters can alter the outcome of experiments conducted with MPS. With many MPS devices being developed, we have outlined some established standards for designing MPS devices and described techniques to validate the devices. A physiologically realistic mimic of the human body can help determine the dose response and toxicity effects of a new drug candidate with higher predictive power.