Quantitative descriptions of natural disturbance regimes are lacking for temperate forest regions in Europe, primarily because a long history of intensive land-use has been the overriding driver of forest structure and composition across the region. The following contribution is the first attempt to comprehensively describe the natural disturbance regime of the dominant forest communities in the Dinaric Mountain range, with an emphasis on the range of natural variability of regime components for the main disturbance agents. Compared to other forest regions in Europe, the mountain range has a history of less intensive forest exploitation and provides a suitable record of natural disturbance processes. Our synthesis is based on multiple types of evidence, including meteorological information, historical documentation, evidence from old-growth remnants, and salvage logging data from National forest inventories. Taken together, the results show that no single disturbance agent dominates the regime in the dominant forest types (i.e. beech and mixed beech-fir forests), and any given agent exhibits remarkable variation in terms of severity and spatial extent both within and among individual disturbance events. Thunderstorm winds cause the most severe damage (i.e. near stand replacement), but blowdown patches are typically limited to stand-scales (e.g. 10s of ha). Ice storms and heavy snow typically cause intermediate severity damage and affect much larger areas (e.g. 100s of km2). A notable exception was the 2014 ice storm, which was nearly an order of magnitude larger and more severe than any other event recorded in the synthesis. Severe and prolonged periods of drought have occurred several times over the past century, and along with secondary insect damage (e.g. bark beetles), have caused episodes of forest decline. Overall, our synthesis indicates that on top of the background of relatively continuous gap dynamics, stand-scale intermediate severity events are an important part of the regime ; these events likely have rotation periods that are less than the lifespan of a tree cohort (e.g. several centuries) and create canopy openings large enough to alter successional trajectories.