In this paper, oil recovery and permeability reduction of a tight sandstone reservoir in immiscible and miscible CO2 flooding processes are experimentally studied. First, a series of saturation tests are conducted to determine the onset pressure of asphaltene precipitation from a light crude oil CO2 system. Second, the vanishing interfacial tension (VIT) technique is applied to determine the minimum miscibility pressure (MMP) between the light crude oil and CO2. Third, a total of nine CO2 coreflood tests under immiscible and miscible conditions are performed through the so-called dry, secondary, and tertiary oil recovery processes, respectively. It is found that the onset pressure of asphaltene precipitation is much lower than the MMP. In the CO2 secondary oil recovery process, the coreflood test data show that, when the injection pressure is between the onset pressure of asphaltene precipitation and the MMP, the oil recovery factor is higher but the oil effective permeability reduction is larger at a higher injection pressure in the immiscible CO2 flooding. They both reach almost constant maximum values in the miscible CO2 flooding (P >= MMP). It is also found that, in three different miscible CO2 oil recovery processes, the CO2 tertiary flooding process gives the lowest oil recovery factor but the largest oil effective permeability reduction. This is attributed to the most severe codeposition of asphaltenes and metal carbonates. However, the CO2 dry or secondary flooding process has a significantly higher oil recovery factor but a much smaller oil effective permeability reduction due to asphaltene deposition alone in the former process or codeposition of asphaltenes and metal carbonates in the latter process.