Mountain pine (Pinus uncinata Ramon ex DC) is an important forest species in the Pyrenean Mountain, both in France and Spain [1]. In a project dealing with its use as timber [2], physical and mechanical tests were performed on clear wood specimen. Wood properties are determined both by genetics (species and provenance) and by growth adaptation to the tree's environnement through variations in ring width (RW), basic density (BD), microfibril angle (MFA) and chemical composition of the cell wall matrix. The square of sound speed called here specific modulus (SM) is a good proxy for MFA [3] and fibre saturation point (FSP) can be a proxy of chemical composition of the cell wall. Shrinkage parameters: total radial (RS), tangential (TS), volumetric (VS) as well as shrinkage anisotropy (SA=TS/RS) are important physical properties for timber use. Longitudinal modulus of elasticity (MOE), resistance to compression (CS) and deformability (strain limit before damage beginning in compression) are key mechanical properties both for tree mechanics and timber technology. All these properties are known to vary, even in a given species, often up to two or threefold, mostly due to adaptation to growth condition, and this is a drawback for timber use. This presentation will focus on the links between variations of growth parameters and wood properties with the objective to build regression models of properties.