A fire behaviour model based on the complete physical and multiphase approach has been developed in order to study fuel treatment efficiency at the wildland urban interface. The fire behaviour model is currently running in 2 dimensions (x, z) and requires a complete description of the vegetation in a 25 cm x 25 cm grid. In each elementary cell of this grid, fuel families have to be identified and quantified. The finest fuel families (twigs less than 6 mm and leaves or needles) have to be described in priority, because they are the most important fuel particles for fire behaviour. The main physical, chemical and thermical properties of each fuel family has to be known and the volume fraction enables to quantify its presence in a given cell. Pinus halepensis stands are fire prone communities very common in South Eastern France. The architectural approach was applied to describe Pinus halepensis fine fuel distribution in order to build up inputs for the fire model. Architectural analysis alms at a comprehensive and dynamic understanding of plant growth through the analysis of the major successive morphological events that happen during plant development from germination to death. The plant architectural software AMAPsim developed by Cirad which relies on both qualitative and quantitative tree architecture description and leads to realistic 3D computing trees can be used to complement fuel characterization. From the computerized plant architecture model it is possible to access and extract various physical parameters. These spatialized data then could be used in fire propagation model. This paper focuses on Pinus halepensis fine fuel characterization using AMAP tools. The AMAP methodology is presented as well as the data collected for plant growth and architecture modelling. Results on the main features of Pinus halepensis architecture are presented as well as Aleppo pine simulations of individual plants in order to extract fuel parameters. The first results of fire sim