Recently, underwater wireless sensor networks (UWSNs) have seen increasing popularity owing to their extensive applications in aquatic environments, including monitoring underwater pipelines, detecting pollution and disasters, researching marine life, underwater surveillance, and facilitating military surveillance. In fact, the task of devising an adequate routing algorithm is particularly challenging because of the unique underwater environmental conditions. These challenges include energy constraints, dynamic topology, long propagation delays, bandwidth limitations, mobility, and 3-D deployments. Therefore, this study addresses the aforementioned challenges and proposes RBEER, a rule-based energy-efficient routing protocol for large-scale UWSNs. RBEER works in three steps: the first is the network initialization and network clustering, in which a Fuzzy C-means is utilized to perform the clustering and determine the cluster centers. The second step is using the RISE rule-based classifier to select the optimal cluster head (CH) based on five input parameters to generate the set of rules. The last step is data forwarding, in which data is forwarded through a single-hop intra-cluster path from member nodes to CH nodes, then through a multi-hop inter-cluster path from CH nodes to sink nodes. Extensive simulations and experiments have been conducted to evaluate the performance of the RBEER protocol. The results demonstrate that the RBEER protocol outperforms benchmarks regarding packet delivery ratio, end-to-end delay, and energy consumption.