Spruce wood specimens were surface-silylated according to three different protocols in order to progressively reduce hydrophilicity and, consequently, adhesion to urea-formaldehyde (UF) glue. Compared to the untreated reference, the macroscopic adhesive strength was drastically reduced in silylated specimens. Specimens treated with the most effective silylation method in terms of reduction of hydrophili city showed near zero adhesion to UF glue. Micromechanical characterisation by means of nanoindentation (NI, Berkovich-type probe) revealed that the wood cell wall stiffness and hardness was not signifi cantly affected by silylation. Contrarily, NI experiments (conical indenter tip with 60 ° opening angle) performed directly at the interface between the wood cell wall and the adhesive showed signifi cantly reduced hardness and reduced specifi c work of NI in silylated specimens. It is concluded that the measured correlation between reduced hydrophilicity in silylated specimens and the mechanical strength of the interface is due to reduced adhesion. This allows calculating the specifi c adhesive energy for the system wood cell wall -urea formaldehyde glue from the difference between the specifi c work of adhesion obtained from the unmodifi ed reference and the most effi ciently silylated specimen. The advantage of this new method lies in the position-resolved measurement of qualitative differences in adhesive energy directly at the interface. This is not feasible with macroscopic test methods, which also include effects of surface roughness, cellular adhesive penetration, or grain angle.