Objective To compare the value of the new national criteria (2 major or one major plus 3 minor criteria) with the Duke criteria for diagnosis of infective endocarditis (IE). Methods A total of 205 patients with clinical diagnosis of IE admitted at West China Hospital of Sichuan University were included in this study. Among them, IE was pathologically confirmed in 97 patients. The sensitivities of both criteria for the diagnosis of IE were compared. Results In 205 cases, the same microorganisms were detected twice in blood cultures in 13 cases (8.3%). Vegetations were detected by echocardiography in 183 patients (89.3%). In 97 cases with pathologically confirmed IE, the same microorganisms were detected twice in blood cultures in 6 cases (6.2%). Vegetations were detected by echocardiography in 89 patients (91.8%). IE diagnose was made in 44 (45.5%) and 86 (88.7%, P < 0.05 vs. Duke criteria) out of 97 pathologically confirmed IE patients by the Duke criteria and new national criteria, respectively. The specificities were 100% and 95.7% by Duke and new national criteria, respectively (P > 0.05). Conclusion With the addition of echocardiographic evidence of endocardial involvement and 2 minor criteria as definite diagnostic criteria, the sensitivity of the new national criteria is superior to that of the Duke criteria for diagnosing IE and the specificity for the diagnosis of IE between the two criteria is similar.