Energy Harvesting (EH) has emerged as a signiffcant technology capable of constructing future wireless sensor networks (WSNs). Although a signiffcant body of studies has focused on designing routing algorithms for EH-aided WSN so that the harvested energy can be efffciently used, most of them do not adequately consider energy overffow and waste caused by the limited battery capacity and data buffer. In addition, these works only concentrated on twodimensional (2-D) plane scenario. In practical environment, sensor nodes are generally deployed in three-dimensional (3-D) space. In light of the above limitations, a brand-new biodynamics inspired energy-efffcient routing algorithm used for EH-aided 3-D WSN is proposed (BDRATD for short). Firstly, through fully considering the spatial distribution of nodes (including planar and vertical dimensionality), we design a classiffed channel model adapted to 3-D spatial communication conditions. On this basis, an energy assessment model is is built to determine the energy state of nodes. Next, we establish a buffer queuing model to detect the blocking level of nodes. Combining the above models, we further consider the inffuence of IEEE 802.11b medium access control (MAC) protocol, and design an idle energy prediction mechanism to improve power utilization. With the help of hemodynamics, we provide the path selection strategy of BDRA-TD for EH-aided 3-D WSN. Moreover, we analyze the complexity of our algorithm and demonstrate it has convergent property. Extensive simulation results show that our algorithm performs superior in terms of average end-to-end delay, average energy dissipation, packet success ratio, average residual energy and acceptable performance in average energy variance compared to other algorithms.