The talk will draw some contrasts between what may be called “empirical” models (e.g., regression, neural networks) and mechanistic models (e.g., circuit analogs). Mechanistic models focus on the component parts/sub processes of the phenomenon of interest, and on their interconnections/interactions, which then generate the range of possible system behaviors. Examples will be given of mechanistic models for aspects of cardiovascular, cerebrovascular and respiratory physiology, and application of these models to data obtained in clinical or ambulatory settings.