Abstract: Abstract: In order to obtain the law of the impact of fuel injection pulse width and injection pressure change on the discharge coefficient and understand the causes of this law, this article designed and completed the experiment with commonly used fuel injection pulse width (1 000-2 500 μs) and large scope of injection pressure (40-180 MPa). The momentum method was adopted to measure the fuel injection law in experiment, and the fuel injection quantity in different periods of injection process could be figured out at the same time, with the purpose of calculating the maximum discharge coefficient which corresponded to the duration of needle valve open completely, and the average discharge coefficient in the fuel injection duration in all conditions. The two parameters were used to explore the law of their own variation when the fuel injection pulse width and the injection pressure changed and to find their relationship and difference. With the purpose of obtaining more general results, the ratio of the duration of valve open fully and the injection duration was introduced as a dimensionless parameter τ, through which we could easily know the deviation value of discharge coefficient according to the curve of the mean discharge coefficient deviating from the maximum discharge coefficient. A lot of research has been made on discharge coefficient change in terms of external cause at home and abroad. But most of these studies were based on fluid simulation, or through Nurick simplified model to analyze, or just discussed the impacts of some physical parameters such as nozzle geometry, needle valve lift, Reynolds number and surface roughness inside the nozzle, fuel density and viscosity on discharge coefficient. Just discussing how much effect the conditions would have on the discharge coefficient was not enough, we must study why. This paper introduced the theory of flow loss analysis of nozzle discharge coefficient. Changes of discharge coefficient with the fuel injection pulse width and injection pressure were obtained by experiment. Then we could explain those changes in the law with the method based on the theory of flow loss, which was different from the conventional research methods of discharge coefficient analysis, for our method not only focused attention on how much impact the change of the fuel injection pulse width and injection pressure will cause on the discharge coefficient, but also attempted to explain why. The method confirmed the influence factors of discharge coefficient, namely the local loss and equivalent linear loss. The local loss was associated with shrunk stream and cavitation, and the equivalent linear loss in the direct ratio to a power of -0.25 of flow velocity. The experimental results showed that: When the injection pressure was over 160 MPa and the increment of injection pulse width was 500 μs, the variation of the average discharge coefficient was no more than 5.9%; when the injection duration was greater than 1.78 ms, the injection pulse width had no influence on the maximum discharge coefficient in principle; when the ratio of the duration of valve open fully and the injection duration was about 0.85, the difference rate between the average discharge coefficient and maximum flow coefficient under the condition of injection pressure below 180 MPa was less than 7.5%; when the injection pressure was greater than 140 MPa, the discharge coefficient tended to be stable and was lower than the value corresponding to injection pressure less than 90 MPa. The method that analyzes the effect of fuel injection pulse width and injection pressure on the discharge coefficient from the angle of flow loss can preferably explain the experimental results, which proves that the method is relatively accurate. This paper can provide a reference for the optimization of injection and the further study of discharge coefficient.