This article integrates core and well log data to determine reservoir electrofacies of the Oligo-Miocene Asmari Formation in the western Dezfol Embayment, SW Iran. At the start, an unsupervised neural network was employed based on the selforganizing map (SOM) technique to identify and extract electrofacies groups of Asmari Formation in the Marun oilfield, which is a mixed siliciclastic and carbonate reservoir. Using inexpensive and accessible petrophysical wireline logs, such as gamma ray, sonic, density and neutron, along with calculated reservoir data (water saturation and effective porosity) from 12 reference wells and their correlation with reservoir core data, led to recognition of seven electrofacies including first four class as limestone rocks (EF1–EF4) and others three classes as sandstone rocks (EF5–EF7). Based on the porosity and permeability maps and water saturation at different depth levels, EF4 and EF7 with low porosity and permeability, high water saturation are considered as non-reservoir relating mostly to sedimentary textures of mudstone and argillaceous sandstones, respectively. By contrast, EF1 and EF5 with high values of porosity, permeability and low percent of water saturations characterize the best reservoir quality rocks, and EF2, EF3 and EF6 have medium reservoir quality. EF1 consists of dolomitic skeletal grainstone, in which biomoldic, vuggy and intercrystalline porosity is the dominant pore type, while the intergranular porosity is the major pore type in EF4 as an unconsolidated sand electrofacies. In general, the results obtained in this study indicate a satisfactory agreement between core data and log facies. This correlation allows rockfacies to be classified in the cored wells and predict those facies in the uncored wells. Using the methodology outlined here, it is possible to track reservoir electrofacies from conventional well log data.