The present invention relates to the field of data processing methods, in particular to an early-warning method for a heavy-rainfall type landslide hazard. The operation process of the method comprises the following steps: (1) calculating the probability of occurrence of a landslide for different zones in a range to be detected, and dividing danger-class zones, including an ultralow-danger zone, a low-danger zone, a medium-danger zone, a high-danger zone and an ultrahigh-danger zone; and (2) carrying out a rainfall type landslide hazard early-warning analysis process which comprises: (a) analyzing the possibility of occurrence of the landslide at different rainfalls, wherein the rainfall of 0-99 mm is a low-danger rainfall, the rainfall of 90-150 mm is a medium-danger rainfall, and the rainfall not smaller than 150 mm is a high-danger rainfall; and (b) making early warning for the different zones in accordance with that the possibilities of occurrence of landslides for different zones at different rainfalls are different, wherein the higher the class of early warning, the higher the possibility of occurrence of the landslide is, and the classes of early warning consist of 5-class early warning, 4-class early warning, 3-class early warning, 2-class early warning and 1-class early warning sequentially from high to low.