Studies from North-America indicate that large wood is one of the key factors of pristine stream ecosystems in temperate forested ecoregions, which influences channel morphology across a wide range of spatial and temporal scales, and which has been successfully used in stream restoration. In contrast to North America, the relevance of large wood has long been overlooked in Central Europe. Transferability of the results of North American studies is limited, because land-use pressure is especially high in Central Europe and the natural setting as well as the restoration objectives differ from North America. The main objective of the thesis is to help develop a Central European perspective on the significance of large wood in streams and rivers and its use in stream restoration. Four studies are compiled in this thesis: 1. Potential natural state of Central European streams in respect to the amount and distribution of large wood: As a first step to describe the potential natural state of streams in respect to large wood, the results of nine investigations were summarized, in which the amount of large wood present in Central European reference sections was quantified (n = 34). Although these reference sections are among the most natural stream sections and in a “near-natural” condition according to Central European standards, the volume of large wood is low compared to other temperate forested ecoregions (factor 3). The data reveal that the current large wood standing stock is considerably less than the potential amount of large wood. However, the range of volumes found in the study streams can be regarded as the minimum volume of large wood that should be present in a “near-natural” Central European stream. 2. Influence of large wood (single large fallen trees) on channel morphology: The morphological effect of single large fallen trees, which are increasingly used in stream restoration projects, was investigated in six short channel sections in Central Europe. The results show that the single large fallen trees significantly increased structural diversity at almost all spatial scales, particularly in terms of pool volume and cross-section complexity. Furthermore, large wood increased variability of some cross-section parameters, which is of special importance, because habitat diversity is assumed to increase with cross-section variability. Moreover, some rare habitat types are clearly associated with the large fallen trees. These results indicate that single large fallen trees significantly enhance channel morphology within one to several years along a stretch of several tens of meters. 3. Quantification of the potential use and the simulation of the effects of large wood in restoration projects: The potential use and effects of large wood in Central European streams was quantified to assess, if large wood recruitment (passive restoration) and placement (active restoration) are suitable methods to restore a considerable part of the streams and rivers. Large wood recruitment is suited for only a small percentage of the streams and rivers, but total length of these channel reaches is high compared to the length of the reaches restored so far. The potential use of active restoration is much higher. About one fifth of the streams and rivers investigated can be restored by the placement of large wood, if the land uses pasture and grassland do not restrict the use of large wood. 4. Review of restoration projects in which large wood has been used so far: A mail survey on restoration projects in which large wood has been already used revealed the following aspects: Most stream managers used fixed wood structures to initiate natural channel dynamics. Failure rate of these structures is low, and preliminary monitoring results indicate that the hydromorphological status improved rapidly in most projects. However, there is potential for improvement from an ecological and economical point of view. The size and the potential effect of some wood structures on stream hydraulics and morphology is low and can be increased without inferring with local restrictions. Furthermore, in most of the cases, complex natural shaped wood structures could have been used instead of bare cylindrical logs to benefit from positive side effects. The data on the restoration projects investigated indicate that costs can be markedly reduced and positive side effects are to be expected, if wood structures without additional anchoring are used. Therefore, it is highly recommended to use such soft engineering methods in future projects whenever possible.