Due to the widespread application of wireless sensor networks in fields such as healthcare, the battlefield, etc., security has become a prime concern for transmitting information without any data manipulation. For this concern, we introduce a Three-Level Weighted Trust evaluation-based Grey Wolf Optimization (3LWT-GWO) approach for the effective detection of misbehaving nodes and provide an optimal secure route through trusted nodes for delivering the data securely to the destination. The proposed model is categorized into three phases: (a) trust-based clustering, (b) cluster head selection, and (c) optimal data routing. Initially, the sensors are deployed randomly in the region in which the nodes have the initial same energy. Then the clustering of nodes is performed in the first phase by computing the Overall Trust Score (OTS) for each node based on the factors like direct trust, indirect trust, energy trust, Long-term neighbor Recommendation Trust, authentication trust, and link quality trust. This OTS helps to identify unsafe nodes. After the identification of unsafe nodes, clustering is performed. In the second stage, the weight of each node is calculated based on the residual energy, node distance, and energy. Then the node that has the highest weight is nominated as Cluster Head. Next, optimal routing is performed based on the GWO algorithm by computing the Trust Satisfactory degree, distance, energy, and delay. Based on the estimated route, the packet is delivered from the source node to the destination. The performance of the 3LWT-GWO method delivers better results when compared with the prevailing techniques in terms of energy consumption, throughput, network lifetime, accuracy, detection rate, and delay.