Soil moisture content (SMC) plays a key role in the crop production as it act as a nutrient and serves as solvent for other nutrients such as sodium, potassium, carbon, nitrogen. Soil characteristics can be analyzed using spectral reflectance of the soil in the 0.4-2.5 um domain. 25 soil samples of different textures have been collected from four different areas of Aurangabad city, Maharashtra. Soil spectral reflectances were measured in the laboratory with an ASD (Analytical Spectral Devices) Fieldspec-Pro spectroradiometer in the (0.4–2.5 ?m) spectral domain. Dataset consists of 150 spectral refelctances (six SMC level i.e. 0%, 12%, 16%, 21%, 26% and 30% for each sample). Existing spectral indices like Normalized Soil Moisture Index (NSMI) and Water Index SOIL (WISOIL) are sensitive to water vapor absorption. Our study has been focused on to overcome existing spectral indices limitation. To accomplish this, new spectral index i.e. Normalized Index of NSWIR domain for Smc estimatiOn from Linear regression (NINSOL) has been implemented on the dataset. Semi empirical soil model has been used to estimate SMC as it is robust against soil sample texture. From the results, it is observed that NINSOL operates in 2056 nm and 2263 nm spectral range. Performance comparison has been done among NSMI ,WISOIL and NINSOL. NSMI and WISOIL led to R2 value as 0.79, 0.81 respectively and 6.6, 6.2 RMSE value. NINSOL produces better results than existing indices as R2 value of 0.85 and 5.6 RMSE value.