The definition of restoration targets and the evaluation of restoration success require a comprehensive, ecosystem-based analysis of all successional pathways, which proceed along gradients of land use intensification and continue after implementation of restoration measures. In the presented study, such analysis was applied to fen ecosystems. Study areas were river valleys and eutrophic lakes in The Netherlands and in Schleswig-Holstein, the northernmost state of Germany. At these sites high fen degradation has taken place and restoration strategies such as "recovery of peat-forming systems" and "development of species-rich fen grasslands" have recently been pursued. Based on an indirect successional analysis, we derived characteristic shifts of abiotic (hydrodynamics, soil parameter) and biotic (species composition, phytomass production) ecosystem traits relative to increasing land use intensity. Species richness and nature conservation value (NCV) initially increased and afterwards decreased during the process of land use intensification. This floristic change was accompanied by an initial decrease and subsequent increase in system productivity. Indicators for P-availability showed the same trend and there was evidence that nutrient limitation changed initially from N- (tall sedge reeds) to P- or P/N (co)-limitation (small sedge reeds) and afterwards again to N-limitation (wet and mesic grasslands). The successional analysis documented the abiotic requirements of vegetation types and their characteristic species, and thus, allowed for the use of these ecosystem traits as indicators to evaluate the success of specific restoration targets. For example, the decrease of Scheuchzerio-Caricetea species could be related more directly to reduced light availability than to an increase of the groundwater tables. Furthermore, we calculated threshold values of these traits for the occurrence of target species. For instance, a sustainable establishment of light-demanding mesotrophic species can only be expected if the standing crop value is less than 400 g m ^-2. In a further step, we estimated the restoration success in selected study areas by applying the findings of the successional analysis. Results showed that rewetting measures have to be carried out with caution if both restoration strategies are aspired in one specific area. Moreover, restoration success of both strategies is limited in initially highly degraded areas. Sources for target species are often missing and abiotic conditions such as nutrient levels and flooding periods are often inappropriate. Consequently, expectations for restoration success should be adapted to the realistic development potential of the individual system and a cost-benefit analysis has to be carried out to avoid unnecessary management costs.