[Objective] African swine fever virus (ASFV) can cause high mortality in swine and lead to huge economic losses. Therefore to establish a strict and efficient control system, including the sensitive and accurate diagnosis methods, effective warning mechanism, which avoid the spread of ASF in China. The aim of this study was to develop a novel and high-sensitive droplet digital PCR (ddPCR) method to detect African swine fever virus (ASFV). [Methods] The methods of ASFV real-time quantitative PCR (qPCR) and ddPCR were established and optimal reaction conditions were confirmed based on K205R gene of ASFV. Each method was evaluated for linearity, limit of detection and specificity. The methods were tested in 163 specimens which were collected from domestic or imported clinical sample or serum samples. [Results] The results indicated that the method both had a high degree of linearity (R2≥0.998). The detection limit of ddPCR reached 0.36 copies and was approximately 10 copies/reaction, which was approximately a 10-fold greater sensitivity than qPCR. The cross-reaction was performed with other porcine pathogens, and negative amplification of the cross-reaction assay demonstrated the high specificity of this method. [Conclusion] This high sensitivity and specificity method could be used as an efficient molecular biology tool to diagnose ASFV, which is a reserve technique and very important for prevention of the spread of diseases across borders, and it promotes the development of ddPCR in China.