Composting processes are an interesting way to treat organic wastes and produce quality products that may be used as soil conditioners or organic fertilizers. Through two major phases inducing firstly biodegradable organic matter consumption and then humic substances synthesis, composting leads to obtain stabilized, dryer, sanitized and more homogeneous materials, ready to return to the soil. Nevertheless, composting end product's quality largely depends on the way to formulate the initial mixing and manage reactions occurring during the treatment. As composting is based on numerous interdependent biological, physical, chemical and thermal phenomena, this treatment is quite difficult to understand on the sole experimental way. Then, the aim of this study was to produce a mathematical composting model, which could be used as a predictive tool to understand composting phenomena and optimise the treatment. An in-vessel 300-litre pilot composting treatment was studied. Three steps were considered to set up the composting model. Firstly, biological reactions were studied thanks to a respirometric method. A biological model, based on a Monod type kinetics was validated, and kinetic parameters were estimated. Secondly, the gas hydrodynamics was studied in the composting pilot. By using a gas tracing method, gas retention time distribution was measured and the gas flow pattern was modelled. Gas flow model was then used to propose heat and mass transfers models within the composting pilot. Thirdly, biological and transfers models were compiled in a one dimension composting model. This latter was programmed and solved with a Scilab software. The composting model enables to predict biodegradable organic matter fractions content, temperature, oxygen content and compost water content through the composting mass all along the treatment. First results were obtained with initial experimental data and non-calibrated parameters. Figure 1 shows results for temperature modelling. The model has now to be well calibrated and further validated on complementary experiments. Yet, it may already be considered as a very interesting tool to understand parameters influence on composting treatment result.