Speleoseismological research carried out in the Pollino Range (Calabria, southern Italy), an area of alleged seismic gap in the active extensional belt running along the Southern Apennines, has placed constraints on the recurrence of M > 6 earthquakes, on the expected ground shaking threshold and on definition of seismogenic sources in the region. Radiometric (U-Th, AMS-14 C, and bulk-14 C) dating of before and after deformation layers from collapsed or tilted cave speleothems indicates that six speleoseismic events have occurred in the area during the last~42 ka, with a mean recurrence of~5.6 ka. Based on the in situ measured geometry and laboratory determined mechanical properties of speleothems and using an ad hoc seismogenic source model for northern Calabria, which involved both normal and strike-slip faults, we evaluate the seismic hazard at the cave sites. The numerical models to compute the ground horizontal acceleration threshold for speleothem failure was tested against intact and currently growing stalactites. The inferred age of these stalactites calibrated using established average speleothem growth rates of 0.3-1.2 cm/ka, ranges from~0.7 to~10 ka, with most of them younger than~5.6 ka. Results show that the~0.8-to 1.0-g peak ground acceleration threshold estimated for collapsed speleothems every 5.6 ka was achieved during strong (M > 6) and close (epicentral distance < 12 km) earthquakes. Considering a mean speleoseismic event recurrence of 5.6 ka and that no speleoseismic shaking has occurred in the last 5.5 ka, the probability of occurrence of an M > 6 event in the area in the next few centuries is quite high.Plain Language Summary Seismic shaking can cause collapse or tilt of cave speleothems (stalactites and stalagmites) and dating the calcite layers immediately after and before the deformation provides the age of unknown prehistorical earthquakes. By dating collapsed and tilted speleothems from the Pollino Range (Calabria, southern Italy), we built a catalog of six speleoseismic events that occurred in the last~42 ka in the area. We defined a seismic hazard model for the cave sites using known or newly proposed seismic sources. The model was tested against an estimate of the vulnerability to seismic shaking for several unbroken and currently forming stalactites based on their geometry (slender speleothems are more vulnerable) and their mechanical properties measured in laboratory. Results indicate that a peak ground acceleration threshold of~1 g for speleothem collapse because of strong (M > 6) and close (distance <12 km from the main seismogenic sources) is reached every 5.6 ka and is in agreement with the age of unbroken speleothems (<10 ka, mostly <5.6 ka). Considering a mean speleoseismic event recurrence of 5.6 ka and that no speleoseismic shaking has occurred in the last 5.5 ka, an M > 6 event in the area is pending in the next few centuries.