Petrophysical evaluation from laboratory and well logging data, such as gamma ray, Nuclear Magnetic Resonance (NMR) and resistivity has been used to identify and evaluate oil zones in reservoirs. Gamma ray log is able to predict lithologies along the well, and X-ray Diffraction (XRD) and X-ray Fluorescence (XRF) techniques also provide important lithological laboratory information. The NMR log directly estimates the porosity of the rock. Resistivity log identifies the water region, but cannot provide information about oil saturation. However, the application of unconventional techniques, such as Electron Spin Resonance (ESR), can be of great relevance in the well logging, adding information about lithologies, oil quality and saturation. The ESR technique, which can directly estimate the oil saturation inside the rock, can also provide important information about the origin and formation of rocks. This technique could be an advance for the oil industry, helping to understand the geothermal processes suffered by the rocks and identifying important regions of the reservoir. Thus, in this work, we studied carbonate rock (limestone and dolomite) samples with and without oil, by ESR technique. Firstly, the effectiveness of ESR in the differentiation of carbonates (calcitic and dolomitic) without oil was tested, through the analysis of manganese ion impurity, which promotes the replacement of calcium and magnesium ions in calcites and dolomites. The samples had a typical manganese ESR signature. The hyperfine parameter, A, was homogeneous for both limestone and dolomites. The zero-field split parameter, D, showed variability, as reported in the literature. The studies were carried out at X-band frequency. The analysis by XRF and XRD provided information on the chemical and mineralogical composition of the carbonates, which were added to the ESR data for application in a correlation matrix, by which it was identified that the components CaO and MgO were directly related to the intensity of the ESR signal and the concentration of CaO was inversely related to the concentration of Al2O3 and SiO2. In one limestone rock sample was incorporated oil, with 25° API and moderate biodegradation level, to identify by ESR, the oil presence inside this rock. The technique was able to identify the oil and its saturation in the rock through the detection of free radical signal, g = 1.9951. The innovative application of the ESR technique to identify oil in varying saturation, in carbonate rock samples from the Illinois Basin (USA), was successful. These results demonstrated that this technique in addition to being accurate, non-destructive and relatively low cost, could be used in mud logging operations providing results for lithological characterization, oil identification and its saturation inside the reservoirs.