According to a generally accepted principle, the data describing test conditions should be always sufficiently complete to enable the reader to repeat the experiment. However, it is not quite an exception if a technical report on or a scientific publication of results of grinding tests carried out in lab-scale tumbling mills omits most relevant data (e.g.: total mass of the mill charge) in favour of less important information (e.g,: number and sizes of the grinding media) or of somewhat doubtful data (e.g.: volume fraction occupied by the mill charge in the grinding chamber). Shortcomings of that kind may occur less frequently if the information on the conditions of the grinding tests can be checked for completeness by referring to relationships which comprise the main variables of the net power draw of the mill and of the energy transferred to the mineral load. In addition, relationships of that kind would offer the advantage of compressing the information on test conditions, and of making test results better comparable as well, A basic relationship serving the said purpose is derived from simplified similarity considerations. A nominal coefficient of mill-internal friction is found suitable to normalize the results of net power measurements. The main free variables of grinding tests are combined in an index which is proportional to the specific energy consumption.