Abstract: To address the limitations of traditional single-data-source suppression, such as low modeling accuracy and poor amplitude fidelity for multiple waves at shot points in the ultra-shallow waters (＜30 m depth) of the western Bohai Sea, where strong seabed reflections lead to the development of various multiple wave types, we proposed a joint multiple wave suppression technique in combination of towed-streamer and ocean-bottom cable (OBC) seismic data. Based on the strategy of “stepwise collaboration and iterative optimization”, this method leverages the high signal-to-noise ratio characteristics of towed-streamer data and the wavefield separation advantages of OBC double-check data. First, multiple waves are suppressed in the towed-streamer data using DWD (water-layer correlation multiple wave suppression) and 3D SRME (surface correlation multiple wave suppression) methods, with their reflected wavefields that serve as high-precision primary wave models. These models are then convolved with the downgoing wavefield of OBC data to accurately predict free-surface multiples at shot points. Finally, effective suppression is achieved through curvelet-domain adaptive subtraction. Practical application on 3D data from the Bohai Sea demonstrates that this joint method could effectively resolve the bottleneck of low modeling accuracy for multiples at shot points in shallow waters, which significantly attenuated multiple wave energy and greatly improved primary wave identification accuracy in OBC data. This technology not only provides a high-quality data foundation for subsequent broadband integration and high-precision imaging, but also holds significant practical value and broad applicability for the exploration and development of complex shallow-water hydrocarbon reservoirs.