The objective was to elucidate the fate of biochar in physically- and chemically-defined fractions of organic carbon (OC) in soil after 510days of laboratory incubation. Fresh corn stover (CS) and biochar produced from corn stover at 350°C (CS-350) and 550°C (CS-550) were added to an Alfisol and an Andisol at a dose equivalent to 7.18tC/ha. After 295days, two undisturbed subsamples from each pot were taken: (i) in one, lucerne (Medicago sativa L.) was seeded and (ii) in the other, the incubation was continued without the plants. On days, 0, 295 and 510, soil samples were fractionated into coarse free particulate organic matter (fPOM), fine fPOM, silt+clay and heavy fractions, and analysed for δ 13 C value and OC content. The fractions were fractionated into dichromate-oxidisable and HCl-hydrolysable OC. After 295days, δ 13 C analysis revealed that >64% of the biochar C was recovered in the coarse fPOM fraction and 13–30% in the fine fPOM fraction. The same pattern was observed after 510days, although a significant increase (P<0.05) in the recovery of biochar C in the silt+clay fraction was observed. Treatments, including the presence of both biochar and plants, induced a significant additional accumulation of OC (P<0.05) in the Alfisol and Andisol amended with CS-350 biochar and CS-550 biochar, respectively, compared with the corresponding treatments with plants but no biochar. We argue the need for a modified physicochemical fractionation method for soils amended with biochar. The use of long term incubations in studies of soil OM turnover (including biochar), where no input of fresh detritus is made, is discouraged.