Introduction Short time scale (~days to year) soil incubation experiments have evidenced that mineralisation rate of soil organic carbon could be increased by higher fresh organic matter (FOM) inputs. This process could affect global soil C stocks but its importance has yet to be assessed at decennial or centennial time-scales. Materials and Methods In this study, we analysed soil organic carbon (SOC) data from a 52-years old bare-fallow experiment in Grignon (France) where plots received no organic matter, or only fresh straw or composted straw. Treatments receiving fresh or composted straw showed no significant difference in SOC stocks dynamics over the 52 years, suggesting no long-term impact of priming effect. To go further, we evaluated whether soil organic matter (SOM) mineralisation rates differed between plots with no input and plots with FOM inputs, using simple models of SOC dynamics based on the Hénin-Dupuis formalism. Results Using a model with 3 SOC pools, we showed that estimated mineralisation rates were 3 to 4 times slower for the plots with no input, suggesting an important role of priming effect. However, a 4-pools model with first order kinetics could satisfactorily fit all the data using a same set of parameters. Conclusion Our results did not assess the absence of priming effect on SOC stocks dynamics at decennial timescale, but suggest that priming effect is not necessarily a relevant process to explain long-term SOC dynamics. It would be worthwhile to test our modelling approach on other long-term datasets, in particular from more nitrogen-limited experiments and using other data giving complementary information on mineralisation rates, such as 14C. (Texte intégral)