Abstarct: Porosity and permeability oil reservoirs are very important properties. Permeability shows the ability of rocks in the conduct such as oil, gas or water trough the pore space of reservoirs will determine the porosity of the rock is said to pore volume. The void may be related or unrelated. The determination of porosity and permeability is a crucial in reverve estimation, production and development of oil reservoirs. The conventional methods for permeability and porosity determination are core analysis and well test.These method are however very expensive and time consuming. Furthermore, one or more well in an oil field may have no core sampels. So, there is a need to use a method could appropriately measure the petrophysical properties of reservoir using available well logs. This study attempts to use artificial neural network and Geolog software for prediction of well logs permeability and porosity of a hydrocarbon field from Persian Gulf. The ANN methods contain Early Stoping (ES) back propagation network, Regularization back propagation network approaches. Using ES method correlation coefficients between core data and predicted permeability in wells sp1, sp5, sp6 are 0.95 and 0.92 and in wells sp2 are 0.98 and 0.90 for train and test data respectively. In the case ES method correlation coefficients between core data and predicted porosity in wells sp2 are 0.95 and 0.91 for train and test data respectively. in the case regularization method, the correlation coefficients in wells sp1, sp5, sp6 are 0.95 and 0.91 for train and test data respectively. Using Geolog software method correlation coefficient between permeability Geolog software and core data in wells sp6, sp5, sp2, sp1 are 0.83 and 0.86 and 0.80 and 0.87 respectively and the correlation coefficient between the porosity Geolog software and the core data in the wells sp2 Is 0.79. The results obtained from this investigation showed that the ANN method is superior to Geolog software to predect reservoir permeability and porosity.