A general multiresolution approach is developed and applied to upscaling of multiscale 2D heterogeneous reservoirs. The method uses the wavelet transformations to the original detailed description of the reservoir, with finer resolution introduced in region of potentially high flow rate (high permeability) and coarser, homogenized property descriptions applied throughout the bulk of the model. Wavelet transformations are currently recognized as the most efficient method of data compression. The method is applied to flow problem (steady single-phase flow) and transport problem (miscible displacement process). In this way, pressure is computed on the coarse-grid using upscaled properties. These pressures are then used to compute the pressure at the small scale within each coarse block. Thus, an approximation for the pressure is obtained without ever having to solve the full fine-grid problem, saving CPU time and memory. Finally, velocity field, the most important key component of fluid flow process, is calculated using obtained pressure field. The results of two problems show good agreement with full fine-grid solution.