Milling tool wear identification literature study, some information characteristics was generated by a number of possible milling tool wear. In theory, the characteristics of each of the information under the milling tool wear has a certain probability of occurring. DS evidence theory, belief functions used to express the size of the probability. the milling tool wear objects was identified through multi-sensor testing, then, measured the information features of each sensor belong to belief functions was obtained, then use the DS combination rule for information fusion, information fusion can be milling tool wear characteristics belong to the reliability function, the final determination based on certain criteria for determining status of milling tool wear type and size of tool wear.