Long-term deflection monitoring data of the Gaojia Garden Bridge is comprehensively investigated to master change laws of bridge deflection with temperature and vehicle load, and to primarily determine variation of deflection. Based on the time-series analysis and wavelet analysis of temperature data, vehicle flow and heavy trucks data from the condition supervision system, this paper studies the correlation between defection variation at mid-span with temperature and vehicle load. As shown in the final results, there is an ideal relativity between defection at mid-span andtemperature,while deflection fluctuation is mainly related to heavy trucks, which will offer effective technical support for the Gaojia Garden Bridge as well as resembling continuous rigid-frame bridge condition assessment.