An indicator to evaluate the proportion of exogenous organic matter (EOM) remaining in soils over the long-term after application has been developed. A database was constructed with analytical data corresponding to 83 EOMs, including sludges, composts, animal wastes, mulches, plant materials and fertilizers. The data included results of proximal analysis (soluble, SOL, hemicellulose-, HEM, cellulose-, CEL, and lignin-like, LIC, fractions, in g kg-1 total organic matter) and of carbon (C) mineralization during long-term incubations under laboratory conditions (in g kg-1 exogenous organic C, EOC). The potential residual organic C after EOM application to soil was assessed from the extrapolation of the incubation results. Then, partial least square regression was used to relate EOM characteristics to the proportion of potentially residual organic C previously determined from the incubations. The biochemical fractions of EOM were not predictive enough to develop the indicator. The proportion of organic C mineralized during 3 days of incubation (C3d) was cumulated and appeared to be the most predictive variable of residual organic C. The proposed indicator of residual organic carbon in soils (expressed as g EOC kg-1) was IROC = 445 + 0.5 SOL - 0.2 CEL + 0.7 LIC - 2.3 C3d. The indicator was calculated for the main types of EOM applied to soils. When compared with the few field data of residual C measured in long-term field experiments, the values provided by the indicator seemed to be over-estimated (i.e. EOC degradation could be faster under field conditions than during laboratory incubations).