The importance of miscible displacements in the petroleum industry makes their understanding and quantitative prediction critical in decisions on the applicability of certain recovery techniques. In this study, scaling miscible displacements in porous media was investigated using a general procedure of inspectional analyis. The procedure was used to derive the minimum number of dimensionless scaling groups which govern miscible displacements. It was found that scaling miscible displacements in a two-dimensional, homogeneous, anisotropic vertical cross-section requires the matching of nine dimensionless scaling groups. A numerical sensitivity study of the equations was performed to investigate the effects of some of the scaling groups on the performance of miscible displacements. Through this sensitivity study, it was found that one of the groups is insensitive to the results over all practical values. Hence, the problem can be scaled by only eight dimensionless scaling groups. The prediction of the recovery efficiency for miscible EOR processes can be achieved solely by analyzing these scaling groups. Preliminary results indicate that when the groups are used as inputs to an artificial neural network, the efficiency of the displacement can be accurately predicted.