Compression wood is a common tissue present in the trunk, branches and roots of mechanically stressed coniferous trees. Its main role is to increase the mechanical strength and regain the vertical orientation of a leaning stem. Compression wood is thought to influence the climate signal in different tree-ring measures. Hence trees containing compression wood are mostly excluded from tree-ring studies reconstructing past climate variability. There is a large gap of systematic work testing the potential effect of compression wood on the strength of the climate signal in different tree-ring parameters, especially stable isotope records. Here we test for the first time the effect of compression wood contained in montane Norway spruce (Picea abies L. Karst) on both δ13C and δ18O tree-ring cellulose records by analyzing compression and opposite wood radii from several disturbed trees together with samples from undisturbed reference trees. We selected four trees tilted by geomorphic processes that were felled by wind, and four undisturbed reference trees in the Tatra Mountains, Poland. We qualitatively classified the strength of compression wood using wood cell anatomical characteristics (tracheid shape, cell wall thickness and presence of intercellular spaces). Then we developed tree-ring width, δ13C and δ18O chronologies from the compression wood radii, the opposite radii of the tilted trees, and from the reference radii. We tested the effect of compression wood on tree-ring cellulose δ13C and δ18O variability and on the climate signal strength. We found only minor differences in the means of δ13C and δ18O compression (δ13C -22.81‰, δ18O: 28.29‰), opposite (δ13C -23.02‰; δ18O: 28.05‰) and reference (δ13C: -22.78‰; δ18O: 27.61‰) radii. The statistical relationships between climate variables, δ13C and δ18O remained consistent among all chronologies. Our findings suggest that moderately tilted trees containing compression wood can be used to reconstruct past geomorphic activity, and for stable-isotope based dendroclimatology.