Detection and arrival picking of microseismic events with low signal-to-noise ratios (S/N) are problematic because these events are usually obscured by ambient noise. We have developed an intraevent coherence-based event detection method to address this problem. The innovations of this method include the adaptation of a crosscorrelation, least-squares-based algorithm to achieve better moveout correction for successive record segments and the use of a multichannel semblance coefficient to identify the microseismic events. After finding the events, we adopted a new picker to determine their P- and S-wave arrival times. This picker was developed by combining three aspects of the distinction between seismic signal and ambient noise, namely, the (1) amplitude, (2) polarization, and (3) statistic property differences. We evaluated the performance of the proposed methods using a real data set recorded during an 11-stage hydraulic fracture stimulation. We have determined that, for microseismic event detection, the proposed method has an overall false trigger rate of 12%. As for arrival picking, the average picking error of the new picker is and its standard deviation is . Comparison of the results of different event detection and arrival picking methods versus the S/N of the data demonstrates that the proposed methods are more applicable for detection and arrival picking of low S/N microseismic events.