Wireless sensor networks (WSNs) are composed of a large number of sensor nodes that are self organized. The open nature of WSNs makes them easily exposed to a variety of attacks and brings many security challenges. Furthermore, because of the potentially dynamic behavior of WSNs, there is hardly any infrastructure or centralized control. To enhance the security in WSNs, it is necessary to evaluate the trustworthiness of nodes since some of them may refuse to forward packets for saving their limited resources. In this paper, we propose a fuzzy-based trust management in WSNs. All neighbors of one node are ranked by their trust value. Only those with higher trust values can be chosen to forward packets. Our trust model exploits multiple metrics, rather than a single metric, and converts them into a single numerical ranking. We perform an thorough evaluation of our proposed approach. The results indicate that our trust model can not only identify abnormal nodes’ behaviors, but can also reduce query traffic while improving the reliability of the exchange messages. In the harsh environment where node density is sparse, the slow-moving nodes cannot effectively utilize the encountering opportunities to realize the self-organized identity authentications, and do not have the chance to join the network routing. However, considering most of the communications in opportunistic networks are caused by forwarding operations, there is no need to establish the complete mutual authentications for each conversation.