In this paper, we suggest the use of the chi‐square test for detecting backoff misbehaviour in IEEE 802.11 EDCA networks. A performance evaluation is performed to compare the chi‐square test with two other methods, known in the literature. To perform a suitable comparison, these two methods are extended to support EDCA and the BEB mechanism. We assume a misbehaviour model, which can be easily executed by a selfish user. We show that the chi‐square test outperforms the other methods in terms of the probability of misbehaviour detection and time required to positively identify a misbehaving node.