Detection of Sybil attack in mobile ad hoc network has been a challenging issue in the context of network scalability, limited resource and complexity of the proposed methods. Literature review shows that most of the detection algorithms suffer from the above constraints and could not exhibit their proper efficiency and performance. This paper introduces a new Sybil detection method which utilises network scalability and shows its efficiency within the available resources. In the proposed method, fuzzy inference rule is used as tool to initially isolate those nodes whose behaviours do not conform with the genuine nodes. At the later stage, we employ a trained artificial neural network (ANN) to find out the Sybil nodes from the suspected nodes. The use of fuzzy inference rule helps to avoid complex mathematical computations as this rule uses simple if…then clause based on nodes’ attributes which can be easily extracted from a real network. On the other hand, the performance of ANN does not get affected in a scalable network since its learning efficiency increases with larger data set. The proposed algorithm does not need any extra hardware like antenna or receiver, which may reduce the battery backup. The advantage of this technique is that it can find out any number of Sybil nodes at one go and also minimises the chances of false positive. We have evaluated our scheme by using simulation and result shows a satisfactory detection rate with few false positive.