Modern metabolomic methods have advanced phytochemical research, however this is often limited by the logistical challenges of broad-scale geographic and taxonomic sampling. Herbarium collections offer a potential solution, yet concerns regarding metabolic degradation have limited their use in this research area. This study evaluates the viability of using herbaria for large-scale metabolomics by screening 240 samples of Achillea, representing 80 species covering all sections, and spanning over a century of collection dates. To validate this approach, we compared the metabolomic profiles of 11 species from both living and herbarium sources. Multivariate analysis demonstrated that, while sample source was a significant factor, species identity remained the dominant driver of variation. Stable compounds including sesquiterpene lactones and flavonoids were shown to be well-preserved in herbarium specimens, while chemically labile N-alkylamides showed significant degradation. Genus-wide analysis using 240 herbarium samples revealed significant chemotaxonomic signals at the sectional level in alignment with established phylogenies. PCA revealed that sect. Babounya, which encompasses species from xeric and semi-desert regions, is characterised by high flavonoid content. It also identified a group of outlier species from zthree sections containing rare furanolignan compounds, suggesting potential limitations in current morphological classifications. Within-species chemical variation was primarily driven by sampling location rather than sample age, indicating that sample age does not influence chemistry, but that geography is an important consideration when designing herbarium-based metabolomic studies. Our findings provide a new dimension to herbaria by clearly demonstrating that preserved specimens can yield robust metabolomic data, particularly for stable compound classes, which can be used for a wide range of phytochemical research.