In the analysis of regenerative chatter in face milling it was originally assumed that the whole phenomenon can be translated with the help of a single - tooth model. Here it has been attempted to establish -that a multi-teeth model approaches nearer to the experimental values than the single tooth model.