Climate change has heightened the occurrence of extreme compound events, yet their impacts on crop yield traits and the variations in yield sensitivity among different varieties and management practices remain elusive. Utilizing long-term (1981–2010) wheat trait data of 85 stations administered by the China Meteorological Center (CMC) for various wheat varieties and sowing dates, we employed a linear mixed models to explore the effects of extreme atmospheric temperature and humidity compound events on wheat. Our findings revealed that the yield loss (23.24 ka ha-1 dec-1) was primarily driven by an increasing trend (0.25 ∼ 1.50 day dec-1) in hot-dry days (HDD) during 1981–2010. Additionally, cold-wet days (CWD: -1.75 % per standard unit) and hot-dry days (HDD: -1.90 % per standard unit) showed similar adverse effects on yield. This similarity is primarily due to their comparable negative impacts on grain growth, with reductions in grain weight (-1.82 % for CWD and -1.65 % for HDD) and grain filling rate (-1.56 % for CWD and -1.27 % for HDD) during the heading and maturity stages of wheat. However, CWD dominated the unfavorable effects on grain number (-1.18 % ∼ -0.22 %). Canopy height and ear density were more susceptible to early extreme climatic factors. In addition, yield losses under early maturing varieties and late sowing measures were less sensitive to extreme climates. Our findings emphasize the negative impacts on yield compositions, but different changes in wheat yield under different varieties and sowing caused by CWD and HDD. Our findings could assist modelers in comprehending how extreme weather affects agricultural production, particularly in terms of variations in crop trait responses.