Living wood in the tree performs a muscular action by generating forces at sapwood periphery and residual strains in dead sapwood fibres. Dissymmetric force generation around tree trunk is the motor system allowing movement, posture control and tree reshaping after accidents. Rather young trees are able to restore the verticality of their trunk after accidental rotation of the soil-root system due to wind or landslide, leading to typically curved stems shape. The very high dissymmetry of forces for the motor action is associated with the occurrence of reaction wood on one side of the inclined stem during many successive years. A method to reconstitute this biomechanical history from observations after tree felling, on either green or dry wood, is discussed. A selection of 17 such trees coming from 15 different species (13 different families), tropical or temperate, hardwoods or softwoods, were selected and peripheral residual strains were measured in situ before felling, on 8 position for each stem. Matched wooden rods were sawn and measured for their mechanical and physical properties at green and dry states, allowing the estimation of tree growth stress, i.e., the force created by the living wood. It was possible to build easy-to-use conversion coefficients between growth stress indicator (GSI), measured in situ by the single hole method, and growth strain and growth stress with the knowledge of basic density and green longitudinal elastic modulus. Maturation strain, specific modulus (as a proxy of micro-fibril angle) and longitudinal shrinkage are properties independent from basic density, whose variation among species was very large. On the whole range of compression wood, normal wood and tension wood strong relationships between these 3 properties were observed, but altogether no single model based on cell-wall microfibril angle only could be used. Growth forces are the product of 4 parameters: ring width, basic density, basic specific modulus and maturation strain, all of them being the result of wood formation. Thanks to the wide range of wood types and species, simple and highly significant formulas were obtained for the relationship between basic and dry density, green and dry longitudinal modulus of elasticity, basic and dry specific modulus. To estimate ring width in green state from values in dry state, radial shrinkage needs to be measured afterwards on dry specimens. Maturation strains is less accurately linked to late measurements on dry wood, but longitudinal shrinkage offers a rather good solution for an estimation provided that the wood type (softwood, hardwood with-G layer, hardwood without G-Layer) is known.