Abstract: To solve the problem that conventional logging methods can’t be used to effectively identify low-resistivity reservoirs, taking Toutunhe formation on the eastern Fukang slope of Junggar basin, the paper uses the NMR logging technology to classify and evaluate low-resistivity reservoirs. Based on the data of NMR, cation exchange capacity, thin section and grain size analysis, the paper studies the genesis of the low-resistivity reservoirs. Based on which, NMR T2 spectrum is used to identify oil intervals and aquifers and to calculate porosity, permeability and water saturation. Although the reservoir of Toutunhe formation contains a small amount of mud and clay minerals, the special clay mineral constitution of the reservoir can generate relatively high additional electrical conductivity and high irreducible water content, which results in the resistivity of oil zones generally lower than 6 Ω·m and the oil zones showing a low-resistivity characteristic. According to the shape of NMR T2 spectrum and whether there is a peak spectral at 200 ms, oil interval, water-bearing oil interval, oil-bearing water interval and aquifer can be effectively classified. Taking 3 ms as the cutoff of T2 and combined with the porosity-permeability correlation obtained from core experiments, the porosity and permeability of the low-resistivity reservoirs can be calculated, finally the cation exchange capacity calculated from NMR logging is substituted into a W-S model to determine the water saturation of the low-resistivity reservoir. The qualitative identification and quantitative evaluation results of the method for the low-resistivity reservoirs well matche with the experimental results, showing a good applicability in oil production.