The invention relates to a method and device for predicting group credit. The method comprises the following steps: acquiring selected data, dividing the selected data into first class group data containing individuals of known default situations and second class group data containing individuals of unknown default situations; selecting a corresponding predefined group default probability; gaining the group default probability of the first class group data according to the selected predefined group default probability to calculate a first variable; obtaining a group characteristic variable of the first class group data; training according to the first variable and the group characteristic variable of the first class group data to obtain a group default prediction model; predicting a group default for the second class group data according to the group default prediction model to obtain the group default probability of the second class group data; transforming the group default probability of the second class group data into corresponding group credit scores. With the adoption of the method and the device for predicting the group credit, the group credit is evaluated.