AbstractCarbon exchange between soil and the atmosphere plays an important role in the global C cycle, but partitioning soil CO2 emissions by source (soil organic matter mineralization, crop residue decomposition, and root respiration) is difficult. Our objective was to determine the contribution of decomposing wheat (Triticum aestivum L.) straw to CO2 emissions from a Swinton silt loam (fine‐silty, mixed, mesic Typic Haploboroll) under controlled conditions (constant 20°C). Two types of straw (i.e., fresh straw collected shortly after harvest and standing stubble that had “weathered” in the field for a year) were either incorporated into or placed on the soil surface at a rate equivalent to 2800 kg ha‐1. One set of soils was watered every 2 or 3 d to 90% of field capacity and a second set was allowed to dry (from 90% field capacity) to below the permanent wilting point before watering. Emissions of CO2 were measured every 2 or 3 d with a vented chamber connected to a portable CO2 analyzer. Within 2 d, incorporation of straw increased CO2 flux from 0.3 to ≈ 1.5 µmol CO2m‐2s‐1. Surface straw had significantly increased fluxes, but the effect was small compared with incorporated straw. Straw type had little effect on emissions. Total CO2‐C emitted in 77 d from continuously moist soil was 25 g m‐2 with no added straw, 41 g m‐2 with surface straw, and 73 g m‐2 with incorporated straw (values are averages for two straw types). In all, 38% of incorporated‐straw C and 13% of surface straw C was emitted as CO2. Soil subjected to moist‐dry cycles emitted from 36 to 62% less CO2 than continuously moist soil. The Douglas‐Rickman residue‐decomposition model simulated the amount of straw C remaining in the soil as a function of time. In the absence of straw, CO2‐C emitted was similar to the amount by which light‐fraction C decreased, suggesting that this labile fraction of organic matter was a major source of respired C.