Accurate and complete monitoring data is very essential for the metro constructions safety risk analysis and process forecasting.However, missing data is inevitable in the monitoring process due to complexity of construction field conditions.In this paper, an emerging theory called compressive sensing (CS) ,which is widely used in the fields of information theory, signal processing, and wireless communications, has been utilized to reconstruct incomplete data in Wuhan metro line No.3constructions.The experimental simulation shows thatwhen the loss rate is very high or the observation intervals are large, the CS still can basically recovery the original data.In conclusion, CS is suitable and effective for safety monitoring data recovery in subway construction.