This work aims to address the problem of localization in 3D surface wireless sensor networks. First, it reveals the unique hardness in localization on 3D surface in comparison with the well-studied localization problems in 2D and 3D space, and offers useful insight into the necessary conditions to achieve desired localizability. Second, it formulates the localization problem under a practical setting with estimated link distances (between nearby nodes) and nodal height measurements, and introduces a layered approach to promote the localizability of such 3D surface sensor networks. Crossbow sensor-based experiments and large-scale simulations are carried out to evaluate the performance of the proposed localization algorithm. The numeric results show that it can effectively improve localizable rate and achieve low location errors and computational overhead, with the desired tolerability to measurement errors and high scalability to large-size wireless sensor networks.