A fast, fully automatic technique to identify microseismic multiplets in borehole seismic data is developed. The technique may be applied in real time to either continuous data or detected-event data for a number of three-component receivers and does not require prior information such as P- or S-wave time picks. Peak crosscorrelation coefficients, evaluated in the frequency domain, are used as the basis for identifying microseismic doublets. The peak crosscorrelation coefficient at each receiver is evaluated with a weighted arithmetic average of the normalized correlation coefficients of each component. Each component is weighted by the maximum amplitude of the signal for that component to reduce the effect of noise on the calculations. The weighted average correlations are averaged over all receivers in a time window centered on a fixed lag time. The size of the time window is determined from the dominant period in the signal, and the lag time is the time that maximizes the average correlation coefficient. The technique is applied to a three-component passive seismic data set recorded at the Valhall field, North Sea. A large number of microseismic doublets are identified that can be grouped into multiplets, reducing the total number of absolute event locations by a factor of two. Seven large multiplets reflect the repeated multiple rerupturing (up to 30 times on a single fault) and significant stress release. Two major faults dominate the seismic activity, causing at least one-fourth of the observed events.