Debris flows are among the most frequent and hazardous disasters worldwide. Debris flow hazard prediction is an important and effective means of engineering disaster mitigation, and rainfall threshold is the core issue in debris flow prediction. This study selected the Laomao Mountain debris flow in Dalian as the research object and explored the relationship among the percentage of coarse sand content of soil, rainfall conditions and the critical rainfall values that induce debris flows on the basis of field investigation data, combined with the results of a flume test, soil suction measurement and geomechanical analysis. The new multi-parameter debris flow initiation warning models were obtained through the mathematical regression analysis method. The critical rainfall values of debris flows in this area were calculated by the previous research on the mechanism of hydraulic debris flow initiation (HIMM). Lastly, the multi-parameter debris flow initiation warning models were compared and analyzed with the critical rainfall values obtained using the HIMM method and the rainfall information available in historical rainfall data, and the reliability of the models was verified. The comparison results showed that the new multi-parameter debris flow initiation warning models can effectively modify the traditional intensity–duration model and have certain reliability and practical values. They can provide an effectual scientific basis for future work on the monitoring and prediction of debris flow disasters. © 2021 by the authors. Licensee MDPI, Basel, Switzerland.