Poplar-based agroforestry systems are one of the most important farming systems on the temperate plains of China, but soil respiration in those systems has seldom been reported. In this study, poplar leaf litter and residues of the two main crops (wheat and peanut) grown in the agroforestry system were amended to form different litter mixing treatments in field experiments at two sites located in Jiangsu Province, China. We measured soil respiration and environmental factors in the different treatments. Soil respiration rates were increased by the addition of plant residues but were strongly influenced by residue quality. During the growing season, soil respiration was negatively related with C/N ratio, while positively related with the initial P concentration of residues (P < 0.05). Poplar leaf litter and crop residues showed non-additive effects on soil respiration when they were mixed. Both air and soil temperature at 10 cm depth explained more than 85 % of the variation of soil respiration at both sites with an exponential model. A significant linear relationship between soil respiration and soil water content at 10 cm depth (W-S) was also observed. The percent of variation in soil respiration explained by a model based on air temperature and soil water content was greater than that explained by a model based on temperature alone. Thus, soil respiration in the studied poplar-based agroforestry systems was driven by both temperature and soil water content. Soil respiration was significantly different between the two sites that had different clay content and C/N ratios. Results from this study are important for us to understand how soil respiration responds to litter mixing or is influenced by biophysical factors in poplar-based agroforestry systems.