Carbon dioxide (CO2) flooding has been used as a commercial process for enhanced oil recovery since the 1970s. Significant amounts of residual oil can be recovered by this procedure. The application of carbon dioxide (CO2) to enhance oil recovery can induce precipitation and deposition of asphaltenes. The consequences generally include changes in reservoir wettability, formation damage, and wellbore and downhole facilities plugging, affecting the project budget due to production delays and costly clean up procedures. The amount of gas that can be dissolved in the oil before having asphaltene precipitation is an important parameter to consider in the design of the optimal gas injection scheme. The challenge of predicting asphaltene precipitation onset is attributed mainly to the difficulty of modeling the oil and the asphaltene molecules. As this behavior is very difficult to describe, the problem caused by this precipitation in all sectors of the oil industry is serious and complex. Therefore a simple method to predict the onset of precipitation would greatly benefit the oil exploration area. In this paper a new method is described which enables the detection of the conditions at which the onset of asphaltene deposition occurs. This approach is solely based on monitoring the difference between the asphaltene solubility parameter and the solubility parameter of the solvent containing the non-asphaltene liquid and CO2. To use this method only two parameters are fitted with the saturation pressure: the binary interaction parameters between the lightest and the heaviest components and the density of C-7(+) fraction. The experimental data required for the implementation of the method are the oil composition stops at C-7(+) fraction, the bubble point pressure at the reservoir temperature and the SARA (saturates, aromatics, resins and asphaltenes) analysis. The applicability of this method to predict asphaltene onset precipitation is demonstrated by extensive tests using data from literature. It corroborates the proposed approach which provides reasonably good predictive results. The examples shown not only validate the proposed method but pose an interesting alternative compared to the classic methods of determining asphaltene precipitation. (C) 2013 Elsevier B.V. All rights reserved.