One of the most important reservoir characteristics for building a three-dimensional geological model is porosity. The best way to get it is to measure it in a lab. However, this method is expensive and time consuming. Information on well mapping and brainstorming are local measurements that do not reflect the behavior of the reservoir as a whole. In addition, well mapping information does not cover the entire area of the field, while 3D seismic information covers a larger area. Many experimental approaches have been proposed to predict these reservoir properties, such as multiple linear / nonlinear regression. In this study, support vector regression was used using well drawings, seismic data and its markers for three-dimensional porosity modeling. By the created model, the porosity was obtained in three dimensions. In the training phase, a high correlation coefficient is obtained between the actual porosity and the estimated location of the test well. Therefore, it can be concluded that this is a powerful method for simulating the three-dimensional porosity of hydrocarbon reservoirs.