Despite the importance of understanding the historical dynamics of ecosystem services (ES), littleresearch has focused on a historical, spatially explicit, assessment of ES supply. This research is aimed at understanding the spatial patterns and potential drivers of temporal variations of ES supply. It has assessed associations of ES temporal variations, delineated ES bundles from changes in ES supply over time, and identified potential drivers of ES bundles. Finally, we discuss the potential implications for spatial planning. We reconstructed the spatio-temporal pat-terns of 11 ES supply in the canton of Vaud, Switzerland, between 1979 and 2014. We used Spearman’s rank coefficient, k-means clustering and redundancy analysis to understand the spatial patterns and potential drivers of temporal variations of ES supply. Municipalities were grouped into four clusters based on ES supply changes over four decades. Food production showed the most negative associations with other ES. Regulating ES were not always synergetic and were less related to increases in population density than cultural ES, which were found in low population density municipalities. In general, synergetic ES may not respond to the same potential drivers. Municipalities were able to supply ES at different levels but none showed an increase in all ES. ES can be synergetic in one bundle, but antagonistic in another. Different processes can cause a change in the same ES depending on their supply location. It seems unrealistic to require each municipality to have a multifunctional territory in the current political context.