Abstract：Many hypotheses or mechanical models are used to describe the mechanical properties of rock，in which some parameters are determined by experimental results. The mechanical model should have feasible and wider applications other than the experiment itself. Mathematical equations are used to fit the experimental data if the mechanical models have not been clearly invented. It is better to optimize fitness equations based on the mechanical and dimensional analyses，and to show the rock mechanical properties with the parameters in the equations. The relations between axial bearing capacity and confining pressure of cylindrical specimens with various sizes are introduced；and the differences between mechanical models and fitness equations are analyzed. Some fitness equations in literatures are discussed. The fitness equations should be shown with such a physical fact that the experimental results are not merely determined by the experimental parameters，also influenced by some factors which can not be controlled. The fitting precision can be improved better when more parameters are used in the equation，but excess of fitting precision may make the equation untrue totally.
