Description: Highlights: • We summarized 20 ecosystem services provided by coastal Baltic ecosystems. • Information on how services translate into socio-economic benefits is lacking. • The deep knowledge gap impairs the sustainable management of the benefits. • We propose a framework with four key measures to close this knowledge gap. • Toxins and Nutrients are the most well-document pressures to these services. Abstract: Seagrass meadows, algal forests and mussel beds are widely regarded as foundation species that support communities providing valuable ecosystem services in many coastal regions; however, quantitative evidence of the relationship is scarce. Using the Baltic Sea as a case study, a region of significant socio-economic importance in the northern hemisphere, we systematically synthesized the primary literature and summarized the current knowledge on ecosystem services derived from seagrass, macroalgae, and mussels (see animated video summary of the manuscript: Video abstract). We found 1740 individual ecosystem service records (ESR), 61% of which were related to macroalgae, 26% to mussel beds and 13% to seagrass meadows. The most frequently reported ecosystem services were raw material (533 ESR), habitat provision (262 ESR) and regulation of pollutants (215 ESR). Toxins (356 ESR) and nutrients (302 ESR) were the most well-documented pressures to services provided by coastal ecosystems. Next, we assessed the current state of knowledge as well as knowledge transfer of ecosystem services to policies through natural, social, human and economic dimensions, using a systematic scoring tool, the Eco-GAME matrix. We found good quantitative information about how ecosystems generated the service but almost no knowledge of how they translate into socio-economic benefits (8 out of 657 papers, 1.2%). While we are aware that research on Baltic Sea socio-economic benefits does exist, the link with ecosystems providing the service is mostly missing. To close this knowledge gap, we need a better analytical framework that is capable of directly linking existing quantitative information about ecosystem service generation with human benefit.