Abstract: Affected by the complex geological conditions and the varying meteorological conditions, the small watersheds in the mountainous areas often suffer sudden floods after short-term heavy rainfall. In addition, the lack of basic hydrological data in mountainous areas makes it difficult to determine early warning indicators for floods. By constructing a distributed hydrological model, considering the influence of soil water content on the flood process under different rainstorm scenarios, we proposed a method for determining the critical rainfall for ungauged mountainous areas and applied it to Baisha River basin in Sichuan Province. The results showed that the soil water content had a significant effect on the flood process. Under the design rainstorm scenarios of 100 a(374.27 mm), 50 a(330.01 mm), 20 a(270.67 mm), 10 a(224.84 mm), the time to reach the alert flow under saturated soil conditions was 0.41, 0.35, 0.55, 0.63 h earlier than the most dry soil conditions respectively. The critical rainfall was 190.02 mm under soil saturation conditions and 207.46 mm under the driest conditions. When the critical rainfall level comes, all people in the range of 30.85 hm2 from Yangliuping to the downstream should evacuate in time. The study can be used for reference in other similar ungauged areas for timely warning of mountainous flood disasters and reducing losses.