On forested paths, dendrogeomorphology has been demonstrated to represent a powerful tool to reconstruct past activity of avalanches with annual resolution and for periods covering past decades to centuries. Here, we present a new semi-quantitative approach for the identification of past snow avalanche events, which relies on the assessment of the number and position of disturbed trees within avalanche path as well as on the intensity of reactions in trees. Based on a statistical evaluation of the approach, we point to the consistency and replicability of the procedure. In a subsequent step, we demonstrate how dendrogeomorpic records can contribute to the specification of expected runout distances and related return periods of extreme events, an indispensable step in avalanche hazard assessment. Based on the reconstructed distribution of runout distances of 25 events and mean event frequencies at two paths in the French Alps, we successfully derive runout values for events with return periods of ≤300 yr. Furthermore, comparison of relations between runout distance and return periods between dendrogeomorphic data and predictions of a locally calibrated statistical-dynamical model show very reasonable agreement. Within the classical intervals used in hazard zoning (i.e. 10-300 yr), mean and mean square errors amounted to 19.7 m and 28.2 m, respectively, in the first path, and to 23.5 m and 45.8 m, respectively, in the second path. Despite some level of uncertainty related to the limits of both approaches, results suggest that dendrogeomorphic time series can yield valuable information to anticipate future extreme events.