This paper evaluates the improvement benefits of Cardiovascular Disease by a proposed multifactor of Population Attributable Risk (PAR) method. Data from the CVDFACTS project and NHI database are used for empirical analysis. The reduction of medical cost can be estimated when the risk factors are totally dispelled from the colony by the proposed multifactor PAR model. The single-factor PAR model can only estimate that improve benefit of the single risk factor and will repeat estimate the economic benefits of multiple risk factors. It is therefore plausible to conclude that the proposed multifactor PAR model can provide more accurate estimates of economic benefits than that of the single-factor PAR model.