Human running can be modelled as either a spring-mass model or multiple springs in series. A force is required to stretch or compress the spring, and thus stiffness, the variable of interest in this paper, can be calculated from the ratio of this force to the change in spring length. Given the link between force and length change, muscle stiffness and mechanical stiffness have been areas of interest to researchers, clinicians, and strength and conditioning practitioners for many years. This review focuses on mechanical stiffness, and in particular, vertical, leg and joint stiffness, since these are the only stiffness types that have been directly calculated during human running. It has been established that as running velocity increases from slow-to-moderate values, leg stiffness remains constant while both vertical stiffness and joint stiffness increase. However, no studies have calculated vertical, leg or joint stiffness over a range of slow-to-moderate values to maximum values in an athletic population. Therefore, the effects of faster running velocities on stiffness are relatively unexplored. Furthermore, no experimental research has examined the effects of training on vertical, leg or joint stiffness and the subsequent effects on running performance. Various methods of training (Olympic style weightlifting, heavy resistance training, plyometrics, eccentric strength training) have shown to be effective at improving running performance. However, the effects of these training methods on vertical, leg and joint stiffness are unknown. As a result, the true importance of stiffness to running performance remains unexplored, and the best practice for changing stiffness to optimize running performance is speculative at best. It is our hope that a better understanding of stiffness, and the influence of running speed on stiffness, will lead to greater interest and an increase in experimental research in this area.