Soil respiration and its responses to soil moisture and soil temperature under different tillage systems during the period of spring maize growth were investigated in Shouyang Dryland Farming Experimental Station, Shanxi Province, China. The soil respiration rate, soil moisture and soil temperature were determined by dynamic chamber-IRGA method, in the maize field, with three tillage practices, including conventional (CT), reduced (RT), and no-till (NT). The results showed that the changes in soil respiration rates had a single peak curve, and its peak appeared in August The seasonal variations in soil respiration rates under CT, RT and NT were 0.50-4.81, 1.11-5.44 and 0.40-5.89 μmol CO₂ m^-2·s^-1, respectively. The trends in soil respiration between CT and RT were similar, while there was a larger fluctuation in soil respiration with NT. The regression analysis showed that soil respiration had a significant correction with soil moisture or temperature, but little at the peak values of soil respiration. Soil moisture (0-10 cm) could explain 57%-76% of seasonal variations in the soil respiration. The moisture sensitivities of soil respiration were NT>RT>CT. Soil temperature (15 cm) could explain 67%-82% of seasonal variations in the soil respiration. the Q₁₀ was NT (2.47)>RT (2.02)>CT (1.59). The two-factor model y=ae^bTW^c or y=a+bT+cW could better describe the relationship between soil respiration and combination of soil moisture and temperature than the one-factor model. The index-power model of combination of soil moisture and temperature (10-20 cm) y=ae^bTW^c can explain 81%-87% of variations in soil respiration (P<0.01). The sensitivities of three tillage treatments to the combination of soil moisture and temperature were: RT>CT>NT. Soil respiration was affected differently by the hydrothermic factor or by each of the single factor.