Remote sensing technology is a promising means for agricultural drought monitoring in large area, and can continuously obtain long-term time series of crop drought information. Currently, quite a few agricultural drought monitoring indices based on remote sensing technology have been developed from different perspectives. However, different agricultural drought monitoring indices derived from remote sensing have obviously different temporal and spatial adaptability. Selecting the appropriate drought monitoring indices based on different regions and crop growth stages is vital for timely and accurate evaluation of drought impact on crops. It is also important for effective water resource management and drought mitigation. This paper focused on adaptability of the agricultural drought monitoring based on remote sensing, and systematically summarized agricultural drought monitoring indices and their adaptability evaluation methods. Firstly, the satellite data sources and the corresponding sensors that are commonly used for agricultural drought monitoring were summarized. Secondly, the relevant sensitive spectral bands of agricultural drought monitoring parameters were reviewed from 4 aspects, including the atmospheric rainfall, the soil moisture, the change of crop physiology and morphology (such as crop canopy temperature, vegetation water content, and crop morphology and green degree) due to water stress, and the comprehensive parameter information. Thirdly, the adaptations and limitations of various agricultural drought monitoring indices derived from remote sensing were fully discussed which involved 4 aspects: the precipitation monitoring indices based on active and passive radar satellite, such as TRMM (tropical rainfall measuring mission satellite) and SAR (synthetic aperture radar); the soil moisture monitoring indices based on the inversion method of thermal inertia, spectral feature space and microwave remote sensing; the crop water requirement monitoring indices based on the physiological and morphological characteristics of crop mutation; and the comprehensive drought monitoring indices based on integrated reflection of crop drought multiple characteristics. Lastly, the current literature revealed much valuable information about the sensitivity and adaptability evaluation methods for diverse agricultural drought indices derived from remote sensing, which included the spectral feature matching evaluation methods and the multivariate statistical analysis methods considering environmental impact factors of crop growth. Domestic and foreign scholars have achieved great progress on the adaptability of agricultural drought monitoring indices, including the application of multi-source remote sensing data, the sensitive spectral band, the applicability of the indices' own mechanism, and the adaptability evaluation methods. Nevertheless, how to select the most suitable agricultural drought remote sensing monitoring index, according to different regions and crops growth stage, is still a problem. Finally, this paper discussed the highlights, the existing difficulties and the future research trends. 1) The relationship between crop and its growth environment should be considered according to different regions and different crop growth stages; 2) The spectral bands information ought to be increased in order to reduce the ratio of signal to noise in the process of remote sensing data acquisition; 3) The appropriate drought monitoring indices derived from remote sensing should be confirmed, especially in the areas with part of vegetation coverage, because most of the crop growth period is under the condition of partial vegetation fraction; 4) The appropriate spatial and temporal scale for these agricultural drought monitoring indices should be determined according to the research purpose; 5) The study of hyperspectral data and technology in the inversion of precision agriculture drought monitoring indices need to be strengthened; 6) The sensitivity and adaptability of the agricultural drought monitoring indices derived from remote sensing in accordance with its own mechanism is supposed to be further explored.