Despite the recommendations of the 1992 European Habitat Directive regarding the evaluation of the conservation status of ecosystems, information on ecosystem functioning is still poorly integrated in long-term monitoring of European semi-natural grasslands. We investigated how to evaluate several ecosystem properties in relation to services provided by grasslands - plant and litter biomass and chemical composition, soil C and N content – by identifying non-destructive proxies involving little or no additional field sampling effort, or calculated from vegetation species composition. These proxies need to be i) informative, i.e. that can be used to approach one or more ecosystem properties based on linear or non-linear relationship models, and ii) generic, i.e. applicable across a wide range of semi-natural grassland habitats and environmental conditions. Based on a large set of field data collected in a variety of semi-natural grasslands in France, we showed that good proxies for ecosystem properties can be identified when the range of climate or soil moisture conditions was reduced. This suggests that these environmental factors influenced the functional pattern of semi-natural grassland ecosystems. The majority of proxies identified so far, including the community weighted mean of leaf dry matter content (LDMCCWM), the height of the herb layer, and litter thickness, are linked to ecosystem properties through non-linear relationships. They are generic in that they are applicable to several grassland habitat types. The height of the herb layer reflects standing plant biomass relatively well, but different non-linear models have to be used depending on climate and soil moisture conditions. Calculating LDMCCWM from species composition data provides a good approximation of the digestibility and P content of plant biomass. Similarly, litter thickness can be used to approach the digestibility and K content of plant biomass in wet grasslands. No generic proxy was found to approximate litter and soil characteristics across the wide range of grasslands habitats under study. By identifying generic and low-constraint proxies for several ecosystem properties, this work provides guidelines for assessing the functioning of semi-natural grasslands - a key criterion of their conservation status - with limited additional field effort and cost compared to plant community monitoring programs.