Hydraulic fracture treatments are commonly carried out in several space- and time-staggered stages which are designed to stimulate isolated segments of a reservoir (i.e., multistage fracturing). However, repeated hydraulic fracturing of reservoir segments fractured in previous stages is commonly observed. In multistage fracturing this problem is also known as out-of-zone or out-of-pay growth, and we shall call it cross-stage fracturing in this article. Cross-stage fracturing can be detected by automated identification of multiplets, i.e., microseismic events with similar source mechanisms and nearly identical location. We applied multiplet identification to detect cross-stage fracturing on two hydraulic fracture monitoring data sets (Canyon Sand and Barnett Shale formations) one of which is described in this paper. We verified this detection method with the initial microseismic events locations. Cross-stage fracturing was detected only a few minutes after the first microseismic events had been detected and located.