Abstract: The fourth and sixth members of upper Triassic Xujiahe Formation develop structural, structural-lithological and lithological reservoirs with characteristics of low porosity, permeability, resistivity and high irreducible water saturation in Guang'an gasfield of Sichuan basin, southwestern China. In space, relationship of the reservoir porosity and permeability varies greatly, and the reservoir gas and water distribution is complex, resulting in difficulty in gas and water pay recognition. Based on the core calibration of logging data, relationships among lithology, logging characterization, petrophysical property and gas-bearing capacity of the target zones were explored. It can be concluded that the reservoirs are made up of fine, medium- and coarse-grained sandstones with characteristics of low values of gamma ray, density and resistivity and high values of sonic, while the porosity and permeability show no correlation with gas-bearing capacity. The porosity and permeability model were established by considering the sixth and fourth members of Xujiahe Formation separately and subdividing the study area into four blocks, which improves precision of the porosity and permeability interpretation. On the recognition that different gas reservoir types have different gas and water distributions, three methods of high definition induction logging, nuclear-magnetic resonance logging and comprehensive discrimination were jointly applied to distinguish gas and water pays, which increases the accuracy of gas and water pay identification. The application of these methods in the Guang'an gasfield show they are suitable for gas and water pay recognition of structural and structural-lithological reservoirs, whereas not feasible for lithological reservoirs with low structural magnitude whose water saturations of gas pay are transitional with those of water and gas- and water-bearing pay due to incomplete separation of gas and water.