This paper presents a method for assessment of loess collapsibility using the data mining technology.Coefficients of collapsibilities are predicted using the method.The database should be created based on practical engineering,and a prediction model,built with the BP neural network.The predicted loess collapse settlement is compared with the measured loess collapse settlement.Results show that prediction precision of collapse settlement is up to 89% for a specific project case.This indicates that the intelligent method of evaluating loess collapsibility is very useful in engineering