We address an approach for rockfall hazard evaluation where the study area resides below a cliff in an a priori exposure to rockfall hazard, but no historical documentation of rockfall events is available and hence important rockfall hazard parameters like triggering mechanism and recurrence interval are unknown. We study the rockfall hazard for the town of Qiryat Shemona, northern Israel, situated alongside the Dead Sea Transform, at the foot of the Ramim escarpment. Numerous boulders are scattered on the slopes above the town, while pre-town historical aerial photos reveal that boulders had reached the location that is now within town limits. We use field observations and optically stimulated luminescence dating of past rockfall events combined with computer modeling to evaluate the rockfall hazard. For the analysis, we first mapped the rockfall source and final downslope stop sites and compiled the boulder size distribution. We then simulated the possible rockfall trajectories using the field observed data to calibrate the simulation software by comparing simulated and mapped boulder stop sites along selected slopes, while adjusting model input parameters for best fit. The analysis reveals areas of high rockfall hazard at the southwestern quarters of the town and also indicates that in the studied slopes falling blocks would stop where the slope angle decreases below 5-10. Age determination suggests that the rockfalls were triggered by large (M>6) historical earthquakes. Nevertheless, not all large historical earthquakes triggered rockfalls. Considering the size distribution of the past rockfalls in the study area and the recurrence time of large earthquakes in the region, we estimate a probability of less than 5 % to be affected by a destructive rockfall within a 50-year time window.Here we suggest a comprehensive method to evaluate rockfall hazard where only past rockfall evidence exists in the field. We show the importance of integrating spatial and temporal field observations to assess the extent of rockfall hazard, the potential block size distribution and the rockfall recurrence interval.