Spectral feature space has large potential to monitor agricultural drought and it becomes a hotspot in the field of remote sensing drought monitoring approaches. This paper summarizes some spectrum feature space approaches, divides them into three categories: LSTa2NDVI approach, NIRa2Red approach and NIRa2SWIR approach, analyzes deeply their basic principles, mathematical equations and application scope, discusses thoroughly their inner problems and development trends, and also points out their prospect and future work.