Objective To make comparisons of the three models of acute and chronic rheumatic carditis to find out an optimal animal model. Methods AntigenⅠwas a emulsifier mixed by complete freund's adjuvant(CFA) and Group A streptococcus(GAS). AntigenⅡ was mixed by incomplete freund's adjuvant(IFA) and GAS. Female Lewis rats were randomly divided into four groups: A, B, C treatmeat groups were immuned with antigenⅠat the foot pad firstly. Subsequently, rats in group A、B、C were injected antigenⅠ, antigenⅡand activated GAS respectively to make the models of RHD. Rats in control group D were immunized with the same protocol outlined as treatment groups but without GAS. Respectively 7, 12, 24 weeks the rats were sacrificed 24(each group was 6). The blood biochemical item and Hematoxylin-eosin(HE) staining of hearts were detected. Results In group C the mortality was 25%. In group A, the incidence of carditis was the highest. Histopathological manifestations of group A, C was not only revealed acute damage such as inflammatory cell infiltrate as well as group B, but also the Aschofflike cells in the myocardial cells interstitial. But in group A and C there had a great degree of the inflammatory cells infiltration than group B. At 24th week rats in group A detected the rate and degree of valve fibrosis in chronic damage were higher than group B and C. None of rats in group D presented carditis or valvulitis. Conclusion In group A, giving the GAS with continuous stimulation after using the mixed emulsification of CFA and GAS to immune Lewis rats for five times was a appropriate method which could provide an optimal animal model for experimental study of acute and chronic rheumatic heart disease.