Energy efficiency and reliable communication are persistent challenges in underwater Internet of Things (UIoT) sensor networks due to limited resources and variable underwater acoustic conditions. Efficient routing is therefore needed to overcome these challenges and ensure reliable, long-term underwater monitoring. In this paper, an energy-efficient and scalable routing system for underwater networks is proposed, called the Grid-Based Fuzzy Logic Optimal Routing Protocol (GFLORP). The proposed protocol divides the underwater monitoring region into logical blocks and establishes a grid-based layout, enabling organized communication and localized coordination. A fuzzy inference system is used to select cluster heads considering residual energy, link quality, and node degree. To ensure reliable data delivery to the sink, it employs a multi-hop forwarding technique. The simulation results show that GFLORP outperforms the existing state-of-the-art methods in terms of network lifetime, energy consumption, packet delivery ratio, and end-to-end delay. The results show that GFLORP performs well in both sparse and dense deployment scenarios, making it a suitable protocol for dynamic and scalable underwater IoT ecosystems.