Abstract: Practical concepts are set forth and previous studies in the field are summarized. The effects of particle size and grading are considered and it is proposed that the 50 percent size can reasonably be considered the dominant size of flow calculations, provided that the minus 1-in material is less than 30 percent, and preferably less than 10 percent by weight. The profile of through-flow is discussed, and the factors are set forth that must be considered in the determination of the stability of a rockfill during throughflow. The use of model tests to predict mode of failure is described. Details of a specific case of failure-that of Hell Hole - are summarized. Characteristics and conditions which determine the stability of discharge face of an unreinforced rockfill slope are detailed. Factors which effect deep-seated stability are listed, and methods of stabilizing downstream slope are reviewed. Case histories where rockfill dams have experienced throughflow are described. The empirical relationship between flow gradient and velocity in the rockfill voids given by Wilkins and Parkin is accurate enough to provide good flow estimates in most practical problems. It is believed that instability is unlikely to develop if removal of the larger rocks at the toe of the embankment is prevented either by their large size or by their being tied back. Conventional circular arc stability analyses, using pore-pressures estimated from turbulent flownet sketches, can be used to evaluate deep-seated stability.