No-tillage (NT) practices with crop residue mulching are seen as an effective way to accumulate soil carbon (C). The rate of soil C accumulation can be determined by measuring soil C stocks over time (diachronic approach) or along a chronosequence that substitutes spatial history differences for time differences (synchronic approach). The objective of this communication is to compare the diachronic and synchronic approaches for determining the rates of soil C storage under NT in the Cerrado region of Brazil. In 2003 and 2007, soil C stocks (0-20 cm) were determined in three NT fields with 5, 9 and 17 years of NT adoption in 2007 (NT-5, NT-9 and NT-17, respectively), one conventionally tilled field (CT, 30 years of tillage in 2007) and one native Cerrado plot (CE) in Rio Verde (Goia´ s state, Brazil). Soil C accumulation rates were calculated following both the synchronic and diachronic approach. Results from the synchronic approach showed that 30 years of cropping under CT depleted the soil C stock to 34.4 Mg C ha_1, which is a decrease of about 27% of the original levels observed under the native vegetation (CE, 47.1 and 47.3 Mg C ha_1, respectively, in 2003 and 2007). Instead, NT adoption had been accumulating soil C through the evaluated years. Soil C stocks measured under NT areas in 2003 and 2007 were 29.9 and 31.3 Mg C ha_1 (NT-5), 33.4 and 34.4 Mg C ha_1 (NT-9) and 45.8 and 46.4 Mg C ha_1 (NT-17), respectively. Much more moderate rates of soil C accumulation were observed diachronically (0.12-0.28 Mg C ha_1 year_1) than with the synchronic approach (1.33 and 1.27 Mg C ha_1 year_1 in 2003 and 2007, respectively). Soil C stocks under CE between 2003 and 2007 (in the diachronic approach) did not change, indicating that diachronic measurements were accurate. Thus, it appears to be very difficult to eliminate all non-wanted sources of soil C variation (i.e. soil texture, land-use history) analysing the soil C accumulation in a chronosequence (synchronic approach).