Wood is increasingly used in restoration projects to improve the hydromorphological and ecological status of streams and rivers. However, despite their growing importance, only a few of these projects are described in the open literature. To aid practitioners, we conducted a postal mail survey to summarize the experiences gained in central Europe and compile data on 50 projects. Our results indicated the potential for improvement from an ecological point of view, as the number and total wood volume, and the median volume of single wood structures placed in the streams per project was low compared with the potential natural state. Moreover, many wood structures were placed nearly parallel to the water flow, reducing their beneficial effect on stream hydraulics and morphology. Restoration success has been monitored in only 58% of the projects. General conclusions drawn include the following. (i) The potential effects of wood placement must be evaluated within a watershed and reach-scale context. (ii) Wood measures are most successful if they mimic natural wood. (iii) Effects of wood structures on stream morphology are strongly dependent on conditions such as stream size and hydrology. (iv) Wood placement has positive effects on several fish species. (v) Most projects revealed a rapid improvement of the hydromorphological status. Most of the wood structures have been fixed, called ' hard engineering'. However, soft engineering methods (use of non-fixed wood structures) are known to result in more natural channel features for individual stream types, sizes, and sites, and are significantly more cost-effective. Synthesis and applications. Large wood has been used successfully in several projects in central Europe, predominantly to increase the general structural complexity using fixed wood structures. Our results recommend the use of less costly soft engineering techniques (non-fixed wood structures), higher amounts of wood, larger wood structures, and improved monitoring programs for future restoration projects comparable with those in this study. We recommend the use of ' passive restoration' methods (restoring the process of wood recruitment on large scales) rather than ' active restoration' (placement of wood structures on a reach scale), as passive restoration avoids the risk of non-natural amounts or diversity of wood loading developing within streams. Local, active placement of wood structures must be considered as an interim measure until passive restoration methods have increased recruitment sufficiently. Journal of Applied Ecology (2007) 44, 1145 20131155doi: 10.1111/j.1365-2664.2007.01401.x