Drought assessment and monitoring using traditional methods rely on rainfall data, which are limited in arid lands and often is very difficult to obtain near real time and costly. In contrast, remote sensing technology is a method for monitoring of large-scale drought.In this research, drought condition was analyzed using drought indices such as TVDI and NDVI from MODIS sensor data for the Yazd-Ardakan plain, Iran. First, relationship between the drought indices with climatic elements were detected. Coefficient of correlation between TVDI and SPI_6 and SPI_12 were 0.68 and 0.71, respectively. Correlation between NDVI and SPI_6 and SPI_12 were 0.49 and 0.51, respectively. Point correlation between TVDI and SPI_6 in 2004 (as a normal year), 2007 (dry) and 2012 (wet year), were 0.64, 0.78 and 0.67 and for the SPI_12 in the above-mentioned years were 0.65, 0.79 and 0.69, respectively. In other word, efficiency of the TVDI in 2007 is better than the other two years. Correlation of NDVI and SPI_6 in 2004, 2007 and 2012, were 0.41, 0.50 and 0.56, respectively. The correlation between NDVI and SPI_12 in 2004, 2007 and 2012, were 0.52, 0.57 and 0.59, respectively. TVDI which takes into account thermal and reflective bands, and soil moisture, is more accurate than the NDVI, which considers only amount of vegetation of the study area. Results showed that the relationship between vegetation and temperature is negative, while, the relationship between vegetation and precipitation is positive. Using of TDVI can compensate defects of the NDVI and used for identifying and monitoring drought.