Remote sensing-based crop mapping using multispectral temporal images is a reliable source of crop status information. Reflectance in red-edge region can be incorporated in vegetation indices for better results as it heavily depends upon chlorophyll content in the leaves. This research work studies the effect of three Sentinel-2 red-edge bands on fuzzy classification of sunflower crop in Shahabad, Haryana, India. Fuzzy set theory was introduced in the image processing for handling the mixed pixel problems. Supervised modified possibilistic c-means (MPCM) classification approach has been adopted for the identification of sunflower fields due to the capability of handling outliers, noises, extraction of single crop and coincident cluster problem. Classification approach was applied on four different modified temporal vegetation indices. The modified vegetation indices are generated by taking different combinations of red and red-edge reflectance bands in a controlled manner with NIR band. The best vegetation index and suitable red-edge band for the discrimination of sunflower crop were determined. Further, optimization of temporal date images to separate mapping of early sown, middle sown and late sown fields was also identified. From the results of this study, it has been proven that for temporal datasets red-edge-based indices are better than the standard indices for distinguishing between different crops while applying the MPCM classification method.