Performing effective medium access control (MAC) encounters great challenges for mobile underwater acoustic networks (UANs), because it suffers from low signal-to-noise ratio, Doppler shift, large communication latency, and so on. Multi-carrier code-division multiple access (MC-COMA) is a promising modulation technique appropriate for solving the above restrictions. In this article, we design a novel MC-COMA-based cross-layer MAC scheme for mobile UANs, called MCI-MAC, to achieve efficient high-concurrency communication. Specifically, this method allocates spread spectrum sequences dynamically on the basis of velocity, propagation distance, data size, and data grade to improve robustness and flexibility of communication. Moreover, we propose an ant-colony-based optimal channel selection algorithm to decrease inter-carrier interference and inter-symbol interference, which concurrently maximizes energy efficiency and network throughput. In addition, we further propose a solution to solve multihop networks' collisions. Simulation results show that MCI-MAC is more effective than the state-of-the-art methods in mobile UANs.