Remote sensing technology is the major method for monitoring temporal and spatial changes of regional agricultural drought, from which both the vegetation index(NDVI) and the surface temperature (Ts)derived can reveal information of soil water content and drought-suffering status of crops via indicating the response of green vegetation to drought intimidation habitat. Nevertheless, there still exist some limitations when only one of the two parameters is used. The two-dimensional feature space based on NDVI and Ts integrates the physiological and ecological connotations of both parameters, and hence can not only indicate the water-heat threat environment when drought occurs but also display the symptom of crops, thus effectively improving the precision and efficiency of agricultural drought monitoring. Based on a detailed description of the principle of applying NDVI-Ts space to the evaluation of agricultural drought, this paper deals with four representative models for drought monitoring, preliminarily analyzes some non-soil-moisture factors affecting the space, and sums up their advantages and disadvantages in application. Some problems worthy of further attention in this field are also discussed.