The Yinggehai Basin, particularly the Haikou nose structural belt in its northern region, is rich in oil and gas resources with significant exploration potential. However, the low-porosity and ultra-low-permeability reservoirs in this area were characterized by complex geological structures, unclear seismic response patterns, and significant challenges in geophysical prediction. Reservoir prediction was conducted based on the geological background of the Haikou nose structural belt in the northern Yinggehai Basin, utilizing available exploration well data. First, sensitive reservoir parameters were determined by integrating seismic data with rock physics analysis. Subsequently, the area’s geophysical characteristics and hydrocarbon distribution patterns were clarified. A multi-parameter collaborative prediction system was established through the innovative combination of waveform-indicated facies-controlled inversion, lithology-fluid probability prediction, and absorption attenuation analysis. This system significantly improved the accuracy of reservoir parameter inversion. The reservoirs’ compressional-wave impedance, shear-wave impedance, and P-wave-to-S-wave velocity ratio were obtained through waveform-indicated facies-controlled inversion. Furthermore, lithofacies, fluid types, and their lateral distribution were quantitatively evaluated by absorption attenuation analysis and lithology-fluid probability prediction. The spatial distribution of favourable reservoirs was thereby determined. Based on the reservoir prediction results, well B1 was deployed to develop the study area effectively. Finally, the measured porosity was compared with the predicted values, and high consistency with well logging and seismic data was demonstrated. The geological reserves in the Yinggehai Basin are expected to be better evaluated through the study, and the prediction accuracy of favourable low-permeability gas reservoirs is anticipated to be improved.