Abstract: The land surface temperature (Ts)Normalized difference vegetation index (NDVI) space based drought monitoring is vulnerable to the soil moisture, vegetation cover and satellite observation state, which makes the dry edge and wet edge of TsNDVI space unstable, so it is likely to bring more errors in the drought monitoring. The paper develops a method named the general TsNDVI space using the multiyear satellite data in the same period to resolve the instability of the Ts NDVI space. Firstly, the primary TsNDVI space based on the satellite data of some single period of some single year is respectively constructed. Then, from every NDVI, the maximum temperature is extracted from the dry edges of the primary spaces of all years to form the temperature of the general Ts-NDVI space dry edge. In the same way, the minimum temperature is extracted from the wet edges of the primary spaces of all years to form the temperature of the general T s-NDVI space wet edge. Finally, by linear fitting, the equations of dry edge and wet edge of the general TsNDVI space are obtained. The composed general Ts-NDVI space can be applied to all years, in which the satellite data are used to compose the general space. With the MODIS 16-day NDVI (MOD13A2) and 8-day Ts (MOD11A2) data employed, the experiment is conducted over Heilongjiang Province, which is one of the main food production areas in China. The Vegetation Temperature Condition Index (VTCI) based on both the general space and the single space is worked out to demonstrate the difference between the primary Ts-NDVI space and the composed general T s-NDVI space. The relative soil moisture in 10 cm and 20 cm depth, precipitation in the satellite observation period and the cumulative precipitation in the last satellite observation periods are used to validate the improvement of the general Ts-NDVI space. The results reveal that the composed general TsNDVI space using multiyear satellite data is clearly triangular in shape with the definite dry edge and wet edge, which are more approximate to the actual boundary of T sNDVI space than those of the primary single space. The improved method greatly enhances the stability of the Ts-NDVI space, and improves the accuracy of monitoring drought by the Ts-NDVI space. The more unstable the dry edge and wet edge of the primary Ts-NDVI space is, the more obvious the improvement of the composed general Ts-NDVI space is. The general space is more sensitive to soil moisture in the shallow soil. The general space is closely related to the recent rainfall as well as the past cumulative rainfall, and is more sensitive to precipitation than before improvement.