With fast economic development, the total length of high speed roads in Shanghai increases dramatically to 815km by the end of 2013. As Shanghai locates at the Yangtze River delta, there is a layer of averagely 300m Quaternary sedimentary at the surface, the basement of the high speed roads are typically of soft soil and geologically weak. Monitoring ground subsidence of the high speed road is very important for the safety and maintaining work. Traditionally, precise leveling is applied for detecting the height changes of road surface. However, it is laborious, expensive, time consuming, and with limited sparse measurement points. InSAR is a new technology for detection of surface changes with high spatial resolution and precision by using satellite images. In this paper, 24 repeated TerraSAR-X images obtained from April 2013 to November 2014 in Shanghai are used to extract ground subsidence along a segment of the high speed road G15 by using the PS-InSAR method. Appropriate multi-look images are applied in the PS-InSAR data processing to enhance the coherence along the high speed roads. A case study was conducted in Shanghai and a maximum of 16mm/year ground subsidence is found at Baojia.