Key message Mechanical properties of small-diameter round timber from hardwood thinnings of common alder (Alnus glutinosa(L.) Gaertn.), European ash (Fraxinus excelsiorL.), European birch (Betula pendulaRoth. andBetula pubescensEhrh.), and sycamore (Acer pseudoplatanusL.) can be evaluated by non-destructive testing on either standing trees or green logs without wood density determination. Velocity differences between acoustic and resonance methods are influenced by tree species and age. Tree diameter improves the estimation of bending strength but not of stiffness. Context There is a need for a reliable, fast, and inexpensive evaluation method to better sort hardwood thinnings according to mechanical properties for use in potential added-value applications. Aims The estimation by non-destructive testing of mechanical properties of round small-diameter timber of four hardwood species (common alder, European ash, European birch, and sycamore). Methods Acoustic velocity was measured in 38 standing trees and resonance velocity was recorded in green logs from these trees. The logs were then dried and tested in bending. Estimation models to predict mechanical properties from non-destructive testing measurements were developed. Results Large differences between velocities from acoustic and resonance techniques were found. Models based on both non-destructive testing velocities together with a species factor are well correlated with bending modulus of elasticity while models including tree diameter are moderately well correlated with bending strength. Inclusion of density in the models does not improve the estimation. Conclusion Models based on acoustic measurements on standing trees or resonance on green logs together with tree species and diameter provide reliable estimates of mechanical properties of round timber from hardwood thinnings. This methodology can be easily used for pre-sorting material in the forest.