A method for identifying undesired telephone calls (Spit/Spam over IP), preferably in a VoIP network, in which the telephone calls coming in for a telephone subscriber, the callee, from at least one calling telephone subscriber, the caller, are subjected to a test or more precisely to a statistical analysis. The method is developed in such a manner that in the framework of the test for incoming calls, the time of the receipt of each call is determined and in each case the probability that the call is undesired is determined, where for the calculation of the probability the time of the receipt of the call, or a parameter dependent thereon, is related to the temporal distribution of previous undesired calls.