Rapid disappearance of crop residue used as soil surface cover in conservation tillage systems reduces the envisaged soil cover benefits. This study was conducted in a conservation tillage experiment, established in 2003 in Nyabeda, western Kenya, to (1) characterize crop residue disappearance during crop growth, and (2) assess termite activity and characteristics of soil (carbon concentration and aggregate stability) in termite-molded sheetings and mound soil. Loss of surface-placed residue in the presence of macrofauna (defined as >1 mm) was up to 83 % in 3.5 months, compared to 33 % in the absence of macrofauna. Overall, residue loss was up to 34 % higher for buried than surface-placed residue. Termite sheetings had, depending on the cropping system, 11-26 and 25-42 % higher (P < 0.01) carbon than bulk soil (0-5 cm depth) and termite mound soil, respectively. Mound soil had 68 % of the soil as water stable macroaggregates (i.e., >250 µm) compared to 57 and 53 % for bulk soil and termite sheetings, respectively. Also, large and small macroaggregates were elevated under conservation tillage compared to conventional-tillage for continuous maize and maize-soybean rotation systems. We conclude that termites can affect soil carbon and its distribution considerably through their sheetings and strategies to supplement residue under ‘sustained attack' by soil fauna are needed if a specific soil cover rate is to be maintained in conservation tillage.