Abstract: As a new progress in oil and gas exploration in the Bohai Oilfield, the large complex sublacustrine fan reservoir has become a new target for exploration. However, owing to the vertical superposition of sub-channels and the cutting and erosion by later channels, the distribution of sand bodies and the relationship between sand bodies are rather complicated. Sublacustrine fan reservoirs are characterized by strong heterogeneity and it is difficult for reservoir characterization. Taking the Kenli 10-A structure of the Bohai Sea as an example, following the principle of “facies-constraining modeling”, the reservoir modeling of sublacustrine fan is carried out in this paper based on various seismic information and highly integrated geological data from drilling, logging and seismic survey. A new modeling technique for sublacustrine fan reservoir is developed by the authors and presented in this paper. The technique is controlled and constrained by the fan reservoir and the three-dimensional sweetness attributes. After several times of modeling, the probability cube of sandstone is obtained and the fine characterization of the channels of the sublacustrine fan completed. Based on the analysis of the connectivity among sand bodies, reservoir prediction is successfully worked out. The research results show that this method can effectively reduce uncertainties and ambiguities in reservoir attribute simulation, improve the accuracy of reservoir prediction, and lay a foundation for later deployment of exploratory wells.