Compound drought and heatwave (CDHW) events, characterized by the simultaneous occurrence of drought and extreme heat, have posed profound socio-economic challenges in recent years, particularly in China. However, global climate models (GCMs) often exhibit significant uncertainty in simulating such extremes under climate changes. In our study, we employed an emergent constraint (EC) approach to improve the reliability of future projections of CDHW events. Using 24 CMIP6 models across four Shared Socioeconomic Pathways, we find statistical positive relationships between historical temperature trends and future changes of CDHW events characteristics (duration, severity, and magnitude). We further combine these relationships with observational datasets to constrain the increase of CDHW events characteristics over mainland China. Here we show that EC reduce the inter-model variance by an average of 35% for duration, 25% for severity, and 21% for magnitude, and the effects get more prominent as period lengthens. Our findings also suggest that projected CDHW events characteristics are slightly lowered than raw simulations, offering valuable information to support risk management and adaptation strategies.