Stump harvesting after clear-felling may be detrimental for species’ landscape-scale persistence. Sensitivity to it is most likely due to reductions in habitat density, resulting in a non-linear relationship between population size and total habitat amount (i.e. fragmentation effects). Here we summarize theoretical expectations and empirical findings on fragmentation effects in stumps and other types of dead wood in disturbed forest habitats and draw conclusions about the consequences of stump harvesting. Within disturbed patches, some negative fragmentation effects have been observed for beetles and lichens, but most studies have found a linear relationship between habitat amount and population size. At landscape scale, evidence of fragmentation effects in disturbed forest habitats has been detected in some published investigations, but none of them focused on stumps. Thus, although organisms associated with disturbed forest habitats are assumed to be strong dispersers, they may still be sensitive to habitat fragmentation. We conclude that stump harvesting at a moderate level is not likely to increase risks for landscape-scale species extinctions markedly, despite clearly negative local effects. However, due to large uncertainties, adaptive management with monitoring of dead wood-dependent species should be applied wherever stump harvesting is more extensive.