Abstract: Abstract: Canal leakage is the main factor affecting the efficiency of irrigation water. In this paper, the main factors of canal leakage (referring to infiltration) were analyzed and numerical analysis was carried out. In order to obtain the main factors that affected canal leakage, the indoor leakage simulation experiment was carried out. The experimental site was located in Northeast Agricultural University, Harbin(126°45′32″E、45°44′41″N). The test case was made of Plexiglas material, with the size 60 cm ×20 cm×110 cm (length×width×height). Mariotte flask was used to ensure water supply and its diameter was 20 cm and the height was 80 cm. A total of 15 treatments with 3 replicates were conducted. During the experiment, the cumulative infiltration and wet front migration distance of each experiment were recorded. In addition, the soil water characteristic curve of the soil samples was tested and the relevant parameters in the VG model were fitted. The parameters of the fitting and the saturated hydraulic conductivity measured by the ring knife method were input into the HYDRUS-2D software to solve the soil water movement model under the experimental conditions, and the simulate values were obtained. The error analysis of the simulated and measured values was carried out. The root mean square error and the determination coefficient of each experiment were calculated. The results showed that the model was reasonable with the root mean square error of 0.43-4.99 L and the R2 higher than 0.9, and the model could simulate the soil water movement process under the experimental conditions. Based on the measured and simulated values, the influence of the 3 factors on the canal leakage and the influence of these factors on the infiltration process were analyzed. The influence of the slope coefficient on the infiltration process was weak and its influence on the infiltration process could be ignored. Two-way analysis of variance was performed in order to analyze the influence of canal base width and infiltration head and its interaction on cumulative infiltration and wetting front migration distance. The results showed that the statistic value of the canal base width was 55.62 for the cumulative infiltration, 59.46 for the horizontal distance of wet front migration, and 7.42 for the vertical distance. All the 3 statistic values were greater than the critical value of 4.78 under the significance level of 0.01. It indicated that the canal base width had a significant impact on the cumulative infiltration and the wetting front migration distance. The statistic value of the infiltration head at the cumulative infiltration was 2.612, the statistic value of the horizontal distance of wet front migration was 19.14, and the vertical distance was 3.00. The infiltration head had a significant effect on the horizontal distance. The statistic values of the effects of canal width and infiltration head on the cumulative infiltration and the horizontal and vertical distance were 0.30, 0.15 and 0.004, respectively. They were less than the critical value of 3.47, indicating that there was no significant interaction between the canal width and infiltration head (P>0.05). The results of the study can provide technical support for improving the utilization efficiency of irrigation water from the aspect of canal seepage control.