Objective To evaluate the effect of treating Ebstein ’s anomaly w ith tricuspid valve replacement (TVR), and analyze the indications and personali ze the surgical methods of TVR. Methods 35 TVR in 32 cases of Eb stein’s anomaly not amenable to repair from May 1993 to June 2007 in An zhen Hospital were retrospectively. The mortality and complication incidence wer e estimat ed and the KaplanMeier survival curves were constructed to estimate actual sur vival and freedom from thrombus embolism rate, and comparative studies were unde rtaken to e valuate the clinical effect between implanting the prosthesis in the original tr icu spid ring or above the coronary sinus, and between fold the atrialized ventricle or not, and between preserve the autologous tricuspid or not. The univariate an alysis and logistic regression were used to explore the early and late death ris k factors. Results There were 2 early hospital death (6.25%). Followup of 30 patients who survived 30 days ra nged up to 14 years (4.2±3.5years)，there were 3 TVR reoperation due to the pro sthesis deterioration. There were 2 late deaths（6.67%）, and the actuarial surv i val at 5,10 year was 92.1%±4.6%, 86.5%±5.8% respectively in KaplanMeier cu rve. There were 2 events of thrombus embolism that were cured through thromboly tic therapy, thus the freedom from thrombus embolism at 5 and 10 yers reached 91 . 3%±4.5%,82.3%±4.2% respectively. 24 late survivors were in New York Heart Ass ociation(NYHA) functional cl ass I, and 4 in NYHA functional class II. The incidence of atria ventricular nod e rhythm in patients with the prosthesis implanted in original tricuspid ring wa s statistically higher than that of prosthesis implanted above the coronary sinus,and the incidence of paradoxical ventricular septal motion in patients wi thout folding the atrialized ventricle was statistically higher that of with fol ding the atrialized ventricle, but the NYHA functional class was not affected by these factors.The univariate analysis and logistic regres sion revealed that preoperative functional right ventricle/atrialized ventriclelt; 1 and ascites were the independent risk factor for the early death (P=0.023, 0.025), whereas preoperative ascites and edema in lower extremities were the in dependent risk factors for late death (P=0.026, 0.019). Conclusion TVR is a good therapeutically option for Ebstein’s anomaly si nce the operative risk is low, the functional status improved in all patients an d the durability of prosthesis in tricuspid position has been good. In order to maximize the effect of the TVR in Ebstein’s anomaly, personalized procedure sho uld be adopted in deciding whether to fold the atrialized ventricle or not, to p reserve the autologous tricuspid valve or not, and to implant the prosthesis in the original tricuspid ring or above the coronary sinus.