The active fraction of soil organic carbon is an important component of soil health and often is quickly assessed as the pulse of CO.sub.2 released by re-wetting dried soils in short-term (24-72 h) assays. However, soils can lose carbon (C) as they dry and, if soil samples vary in moisture content at sampling, differential C loss during the pre-assay dry-down period may complicate the assay's interpretations. We examined the impact of pre-assay CO.sub.2 loss in a long-cultivated agricultural soil at initial moisture contents of 30 %, 50 %, and 70 % water-filled pore space (WFPS). We found that 50 % and 70 % WFPS treatments lost more C during drying than did those in the 30 % WFPS treatment and that dry-down losses led to a 26 %-32 % underestimate of their CO.sub.2 pulses. We developed a soil-specific correction factor to account for these initial soil moisture effects. Future C mineralization studies may benefit from similar corrections.