Plant nutrition indexes are important tools to estimate forage nutritional status and adjust fertilisation to plants requirements. However, these indexes have initially been developed for grasses and may be ill-suited to grasslands with relatively high abundance of forbs. To test the validity of nutrition indexes (NI) for forb-rich grasslands, we set up an experiment involving nine intensively managed permanent grasslands across Switzerland with contrasted plant composition and soil fertility. e three first harvests of the year 2018 were sorted into grasses, forbs and legumes. Forb abundance in plots ranged from 2 to 74% while legumes represented less than 5% on average. We measured nitrogen (N), phosphorus (P) and potassium (K) content of bulk forage and grasses, and calculated the respective nutrition indexes (NNI, PNI, KNI) separately for both. e PNI and KNI calculated for bulk forage and grasses were slightly different and differences varied among harvests. In addition, we found a strong discrepancy between grasses and forage NNI (up to 70% difference), which was independent of the harvest but influenced by forb abundance. Overall, our findings suggest an increasing underestimation of N fertiliser needs of grasses with an increasing abundance of forbs when using bulk NNI.