Objective To evaluate the significance of clinical grading methods in acute pulmonary embolism (APE). Methods Clinical data of 259 patients suspected APE were retrospectively analyzed. The clinical probability was classified into low, intermediate and high grade by the Geneva score, the Wells score and the SYSU score. The result was contrasted with gold standard. Results Through the three, methods, pa-tients were classified into low pmbability (43.9%-52.5%), intermediate probability (38.0%-42.1%) and high probability (9.5%-14.0%), and the actual frequencies of APE in each category were 6.2%-14.4% in low probability, 65.9%-76.2% in intermediate probability, 88.5%-90.5% in high probability. The SYSU score had the lowest rate of missed diagnosis in low probability (P<0.05 ).The Geneva score was the most accurate in predicting the intermediate probability (P<0.05). But in high probahility, three prediction rules had no significant difference (P>0.05). Combined with D--dimer test, the rote of missed diagnosis in low probability can be lowered. Conclusions The clinical grading methods can predict the clinical probability of APE. It exists similar accuracy, but has different scope of application. Clinical doctor should choose the ap-propriate grading methods in different patients.