1. The importance of wood decay for the global carbon and nutrient cycles is widely recognized. However, relatively little is known about bark decay dynamics, even though bark represents up to 25% of stem dry mass. Moreover, bark presence versus absence can significantly alter wood decay rates. Therefore, it really matters for the fate of carbon whether variation in bark and wood decay rates is coordinated across tree species. 2. Answering this question requires advances in methodology to measure both bark and wood mass loss accurately. Decay rates of large logs in the field are often quantified as loss in tissue density, in which case volume depletions of bark and wood can give large underestimations. 3. To quantify the real decay rates, we assessed bark mass loss per stem surface area and wood mass loss based on volume-corrected density loss. We further defined the range of actual bark mass loss by considering bark cover loss. Then, we tested the correlation between bark and wood mass loss across 20 temperate tree species during 4 years of decomposition. 4. The area-based method generally showed more than 3-fold higher bark mass loss than the density-based method (even higher if considering bark cover loss), and volume-corrected wood mass losses were 1.08-1.12 times higher than density-based mass loss. The deviation of bark mass loss between the two methods was higher for tree species with thicker inner bark. Bark generally decomposed twice as fast as wood across species, and faster decaying bark came with faster decaying wood (R2=0.26, P=0.006). 5. We strongly suggest using corrected volume when assessing wood mass loss especially for the species with faster decomposable sapwood and all the wood at advanced decay stages. Further studies of coarse stem decomposition should consider trait “afterlife” effects of inner bark and estimate fraction of stem bark cover to obtain more accurate decay rates. 6. Our new method should benefit our understanding of the in situ dynamics of woody debris decay and monitoring research in different forest ecosystems worldwide, and should aid meta-analyses across diverse studies.