Artificial structures, serving as the solid matrix of anisotropic porous media and satisfying the requirement needed for flow visualization, were constructed with the perforated Polypropylene plates in order to improve the understanding of transport phenomena occurring in anisotropic porous media. This paper reports the regressed correlations of the experimental pressure gradient and filtration velocity data of three anisotropic and one isotropic porous media measured along two mutually orthogonal directions, which correspond to the principal axes of the permeability tensor, for the filtration velocities ranging from 0.2 to 12 mm/s with water as the fluid. To reflect the observed data, the regression equation with two types of deviations was formulated, in which the pressure gradient is represented by the sum of the linear and nonlinear terms of the filtration velocity. The physical model developed for the linear term assumes the solid matrix as repeated circular orifices when the filtration velocity approaches zero. The exponent of the filtration velocity in the nonlinear term was determined to be that of the Forchheimer extension. Also, four models for the coefficient of the nonlinear term were examined and the results were compared. The distribution of the residuals (the differences between the observed and the correlated values) validated the suggested regression procedure and the resulting correlations.