The knowledge in the field has been reviewed by literature studies, and using experience from personal research in sublevel caving. There is a lot of material regarding gravity flow of fine grained materials in bins, but the theory of flow for these materials are normally not applicable for coarse blasted rock. This study was therefore concentrated towards coarse rock fragments created after blasting in orebodies. The goal for this study was to make a report of known knowledge, and identify what we have to learn more about. The Bergmark-Roos theory, which was a paradigm shift in the science of gravity flow, is today the best mathematical theory for calculation of the drawbody for a homogenous material or when we wish to simulate the mean drawbody for a large number of draws. The Bergmark-Roos theory must however be improved, so it can take into account the flow pipe effect. When this is done it will be possible to construct a 3D computer model for calculation of ore extraction versus amount of loaded ore and waste. It is also necessary to find the 3D shape of the flow cone to be able to calculate the 3D ore rest that normally develops on footwalls. Calculations should be done for the draw on several sublevels including determination of the ore losses and temporary ore rests. First when this computer model is developed optimisation of ore recovery and waste rock dilution can start for a given orebody. The author points out the importance for international co-operation and standardisation of methods for determination of material properties. More field tests and model blasting tests should also be done. When enough knowledge is gathered, a handbook of the optimisation of sublevel and block caving should be written.