A new method is presented here for petrophysical interpretation of heterogeneous carbonates using well logs and core data. Developing this new method was necessary because conventional evaluation methods tend to yield inaccurate predictions of pore connectivity and permeability in the studied field. Difficulties in the petrophysical evaluation of this field are related to shoulder-bed effects, presence of non-connected porosity, rock layers that are thinner than the vertical resolution of well-logging tools, and the effect of oil-base mud (OBM) invasion in the measurements. These problems give rise to uncommon measurements and rock properties, such as: (a) reservoir units contained within thinly bedded and laminated sequences, (b) very high apparent resistivity readings in the oil-bearing zone, (c) separation of apparent resistivity logs with different depths of investigation, (d) complex unimodal and bimodal transverse relaxation distributions of nuclear magnetic resonance (NMR) measurements, (e) reservoir units having total porosity of 0.02 to 0.26 and permeability between 0.001mD to 4.2D, (f) significant differences between total and sonic porosity, and (g) low and constant gamma-ray values. The interpretation method introduced in this thesis is based on the detection of layer boundaries and rock types from high-resolution well logs and on the estimation of layer-by-layer properties using numerical simulation of resistivity, nuclear, and NMR logs. Layer properties were iteratively adjusted until the available well logs were reproduced by numerical simulations. This method honors the reservoir geology and physics of the measurements while adjusting the layer properties; it reduces shoulder-bed effects on well logs, especially across thinly bedded and laminated sequences, thereby yielding improved estimates of interconnected porosity and permeability in rocks that have null mobile water saturation and that were invaded with OBM. Additionally, dynamic simulations of OBM invasion in free-water depth intervals were necessary to estimate permeability. It is found that NMR transverse relaxation measurements are effective for determining rock and fluid properties but are unreliable in the accurate calculation of porosity and permeability in thinly bedded and highly laminated depth sections. In addition, this thesis shows that low resistivity values are associated with the presence of microporosity, and high resistivity values are associated with the presence of interconnected and vuggy porosity. In some layers, a fraction of the vuggy porosity is associated with isolated pores, which does not contribute to fluid flow. An integrated evaluation using multiple measurements, including sonic logs, is therefore necessary to detect isolated porosity. After the correction and simulation, results show, on average, a 34% improvement between estimated and core-measured permeability. Closer agreement was not possible because of limitations in tool resolution and difficulty in obtaining a precise depth match between core and well-log measurements. / text