Soil pH and calcium carbonate contents are often hypothesized to be important factors controlling organic matter turnover in agricultural soils. The aim of this study was to differentiate the effects of soil pH from those related to carbonate equilibrium on C and N dynamics. The relative contributions of organic and inorganic carbon in the CO2 produced during laboratory incubations were assessed. Five agricultural soils were compared: calcareous (74% CaCO3) loess (0.2% CaCO3) and an acidic soil which had received different rates of lime 20 years ago (0, 18 or 50 t ha(-1)). Soil aggregates were incubated with or without rape residues under aerobic conditions for 91 days at 15 degrees C. The C and N mineralized, soil pH, O-2 consumption and respiratory quotient (RQ = Delta CO2/Delta O-2) were monitored, as well as the delta C-13 composition of the evolved CO2 to determine its origin (mineral or organic). Results showed that in non-amended soils, the cumulative CO2 produced was significantly greater in the limed soil with a pH > 7 than in the same soil with less or no lime added, whereas there was no difference in N mineralization or in O-2 consumption kinetics. We found an exponential relationship between RQ values and soil pH, suggesting an excess production of CO2 in alkaline soils. This CO2 excess was not related to changes in substrate utilization by the microbial biomass but rather to carbonates equilibrium. The delta C-13 signatures confirmed that the CO2 produced in soils with pH > 7 originated from both organic and mineral sources. The contribution of soil carbonates to CO2 production led to an overestimation of organic C mineralization (up to 35%), the extent of which depended on the nature of soil carbonates but not on the amount. The actual C mineralization (derived from organic C) was similar in limed and unlimed soil. The amount of C mineralized in the residue-amended soils was ten times greater than in the basal soil, thus masking the soil carbonate contribution. Residue decomposition resulted in a significant increase in soil pH in all soils. This increase is attributed to the alkalinity and/or decarboxylation of organic anions in the plant residue and/or to the immobilization of nitrate by the microbial biomass and the corresponding release of hydroxyl ions. A theoretical composition (C, O, H, N) of residue and soil organic matter is proposed to explain the RQ measured. It emphasizes the need to take microbial biomass metabolism, O-2 consumption due to nitrification and carbon assimilation yield into account when interpreting RQ data.