Correct and accurate determination of the three-dimensional porosity model is one of the necessary steps to reduce the risk and cost of drilling and operation of reservoirs. Regarding the high outlay needed for exploratory drilling, oil companies prefer to use alternative techniques to estimate the porosity model. This study was performed using the information of two wells in one of the southwestern fields of Iran and with three methods of limited band, sparse spike and model-based inversions with the aim of estimating the acoustic impedance as an important attribute that has a high sensitivity to porosity. Here, the model-based inversion gave the best estimation with a 99% correlation with acoustic impedance at the wells. Then, implementing two methods of multi-attributes regression analysis with an error 3.3% and a probabilistic neural network with an error 2.9% led to estimating the porosity model in the entire reservoir.