Stream restoration in Finland has traditionally focused on larger rivers, aiming to enhance the well-being of salmonid fishes. The key strategy has been to increase in-stream habitat heterogeneity, usually with at least some success. Most restoration projects have not targeted biodiversity and stream ecosystem health and it is therefore not surprising that any biodiversity effects have been modest at best. More recently, the focus of stream restoration has shifted to upper parts of the river network, headwater streams. Forestry, particularly the massive transformation of peatlands to commercial forests via drainage ditching, has been historically intense in Finland, causing nutrient enrichment, altered hydrological regime and streambed sedimentation in the recipient freshwater ecosystems. Restoration of headwater streams therefore largely focuses on mitigating the negative impacts of fine sediments on benthic biodiversity and key ecosystem functions. Removing sediments from the streambed has proven extremely challenging, and most restoration measures therefore strive at (i) reducing any further input of fines (streamside restoration), and (ii) increasing the hydrological retentiveness of a stream by adding stony enhancement structures or large wood (in-stream restoration). In a series of field surveys and mesocosm experiments we have assessed the effectiveness of restoration in enhancing benthic biodiversity and stream ecosystem functioning. Our results show clear positive effects of boulder additions on aquatic bryophytes, with indirect positive effects on the diversity of benthic invertebrates (with a time lag). Key ecosystem processes (algal primary production; leaf decomposition) remained largely unchanged. However, the strongest responses to in-stream restoration did not occur in the stream channel but in the adjacent forest. Diversity of riparian plants did not change much but their guild and community composition did, due mainly to increased hydrological retentiveness in reaches with added wood. Log additions helped restore the natural flow regime, benefiting species that tolerate being submerged during floods. Our study showed the intimate linkages between the stream and the streamside forest. Traditionally, management and conservation of streams and riparian forests have followed separate paths. However, as restoration (or any other management action) in one of the two interlinked ecosystems also affects the other one, they should be managed and protected in an integrated effort. Our future studies on stream-forest interactions will focus on (i) effects of stream restoration on the reciprocal energetic subsidies across the stream-forest interface; and (ii) prediction of the effects of altered hydrological regimes on the capacity of variously restored vs. impacted streams to retain terrestrial inputs of organic matter.