The evolution of strut and tie models is reviewed and their characteristics are discussed. From their evolution it is seen that more design effort should be expended on developing a reasonable truss than on refined calculations of nodal stresses and permissible concrete stresses. Strut and tie models are design tools. They allow engineers to put an appropriate amount of material in an appropriate place, and expeditiously demonstrate that there is at least one way for the structure to safely carry the design load. They are not very useful for checking designs. When one is planning the structural analysis, one needs to know if the design will be done on the basis of strut and tie models. Different models may be required for different load cases. Rather than modeling the whole structure, models may be applied to specific regions of a structure. Designs based on strut and tie models may require slightly more reinforcement than designs based on other methods since, in general, the contribution of concrete tensile strength is ignored. Finely tuned empirical equations will use less reinforcement but are restricted by the limits of the test data. Strut and tie models are general rational design tools that do not have such limitations.