Natural or spilled oil in the ocean can interact with marine snow and sediment from riverine sources and form Marine Oil Snow (MOS) aggregates including aggregates consisting of phytoplankton, detritus, and feces. Such aggregates have a fractal structure and can transport oil from the surface layers to greater depths in the ocean, eventually settling on the seafloor. In recent studies of the Deepwater Horizon and IXTOC‐1 oil spills in the Gulf of Mexico, this process was identified as one of the main mechanisms for transporting oil vertically in the water column. We have adapted a stochastic, one‐dimensional numerical model that uses coagulation theory to simulate MOS formation and sinking in the ocean and predict the time evolution of physical properties and spatial distribution of MOS. Here we present the model development, calibration, and validation with measured MOS field data in the Gulf of Mexico during the Deepwater Horizon spill. We use a sensitivity analysis to identify critical parameters, and suggest future model improvements and areas where further experimental investigation is needed to improve our understanding of MOS formation and sedimentation. The model can be used during response and planning activities associated with oil spills in the marine environments.