To evaluate the clinical application of a dosing algorithm by genotypes in prediction of warfarin maintenance dose in Chinese patients with pulmonary thromboembolism. During October 2010 and August 2012, 220 inpatients or outpatients with pulmonary embolism in Beijing Anzhen Hospital, were enrolled by the inclusion criteria. The patients included 86 males and 134 females. The clinical data and blood samples were collected. The fluorescent PCR genotyping method was used to detect the genotypes of vitamin K epoxy compounds reductase complex subunit 1 (VKORC1) and cytochrome P450 enzyme 2C9 (CYP2C9). According to the random number table, the patients were randomly divided into a study group and a control group. In the study group, the first 3 doses of warfarin were prescribed according to the predicted warfarin dose, while in the control group the drug was prescribed according to the dose estimated empirically by clinicians. Warfarin was adjusted until it reached a stable dose according to the INR value, and the following-up lasted for 50 days. AT the end of follow-up, the percentage of patients who obtained a stable dose in the study group and the control group was 82.1% (n = 78) and 66% (n = 64), respectively. The mean time to reach a stable dose in the study group and the control group was (16.8 ± 1.5) and (25.6 ± 1.8) days, and the median time was (11.0 ± 1.0) days and (20.0 ± 2.0) days, the difference between the 2 groups being statistically significant (χ(2) = 18.175, P < 0.001). The incidence of side effects of the study group was lower than that of the control group, and the time to the occurrence of side effects in study group was longer. The average predicted dose of the 142 patients who reached a stable dose was (3.6 ± 0.9) mg/d, and the average effective dose was (3.7 ± 1.3) mg/d, the average predicted dose being lower than the actual dose (0.1 ± 1.2) mg/d, but the difference was not significant(t = -1.202, P > 0.05). The warfarin stable dose prediction algorithm, containing genetic factors and non-heritage factors, can significantly shorten the adjustment time to reach warfarin stable dose, and reduce the incidence of side effects, and is clinically applicable.