One of the key factors in the design of gas injection project is the minimum miscibility pressure (MMP) because local sweep efficiency from gas injection is mainly dependent on the MMP. Slim tube displacement, vanishing interfacial tension, and rising bubble method are three main tests to determine MMP. However these tests are both expensive and time consuming. Our study aims to find a quick and accurate mathematical methodology to determine gas–oil MMP. This study was conducted using 30 wells data sets which have experimental CO2-MMP values. These data sets were corresponding to gas/oil compositional information of 25 blocks. In order to understand miscibility mechanism between CO2 and oil, mass transfer phenomenon between them was firstly displayed in a visible pressure-volume-temperature (PVT) cell apparatus. Different hydrocarbon compositions range effecting on miscibility were presented and reservoir oil components are classified into three pseudo-components for modeling process, which are C2-C15, C1+N2+C16+ and colloid-bitumen. Finally, CO2 density under reservoir temperature, composition coefficient (Xf) and colloid-bitumen content (mass percent) were used to predict pure CO2 MMP, a model with 3 variables was developed. The results of proposed model agree well with experimental data, which indicates that the model was reliable and predictive. This study shows great possibility for future enriched CO2 miscible injection to improve miscibility and enhance oil recovery.