Soybean is an important oil crop in China, and the national focus of soybean production is in Northeast China. In order to achieve high-stable yield, it is crucial to acknowledge the impacts of mean climate and extreme climate indices on soybean yield and yield components. In this study, based on the weather data from 61 counties from 1981 to 2017 in Northeast China, we assessed the impacts of mean climate and extreme climate indices on soybean observed yield and simulated yield. Mean climate include effective growing degree days (GDD10), precipitation (Pre), and solar radiation (SR); extreme climate indices include the number of cool days during seed-filling period (C15), the number of cool days during 15 days before anthesis (C17), the number of hot days (H30), maximum amount of 5 Day accumulated precipitation (P5), and consecutive dry days (CDD)). We used the DSSAT-CROPGRO-Soybean model to identify the main yield components for soybean. The results showed that observed soybean yield reduced by 3.57% due to the collective changes in the eight study climate indices. Increases in GDD10, decreases in Pre, and decreases in SR caused a 3.96%, -3.89%, and - 0.48% change in soybean yield, respectively. Decreases in C15 and C17 led to a 5.36% increase in soybean yield; increases in H30, P5, and CDD caused a 5.75%, 0.30%, and 1.14% reduction in soybean yield, respectively. By comparing the response of observed and simulated soybean yield to climate indices (excluding P5) in the DSSAT-CROPGRO-Soybean model, we identified the key yield components for soybean as the number of pods and seed weight. The negative impacts on the number of pods and seed weight were mainly attributed to changes in Pre and H30 from anthesis to podding and during seed-filling period. Our results could be used to assist the local soybean community adapt to climate change.