Rhynchophorus (curculionid) larvae produce economic damage to ornamental and date palm crops that could be mitigated significantly by early detection and treatment. Acoustic technology enables early detection, but often it is difficult to distinguish insect sounds from background noise containing energy at the resonant frequencies of stiff, fibrous structures in trees and other plants. Tests were conducted with currently available acoustic instrumentation and software to assess the capability of these methods to discriminate curculionid, cerambycid, and buprestid larval sounds from background noise in woody structures. An approach to the discrimination problem is to monitor the temporal patterns of the 3-10-ms sound impulses produced by locomotory and feeding activities. Playback and computer analyses of larval sounds revealed trains of impulses separated by intervals of less than 500 ms that experienced listeners frequently use as indications of potential insect sounds. Further analyses identified a subgroup of trains, denoted as bursts, containing > 6 and < 200 impulses, which occurred frequently when larvae were present but only rarely when larvae were absent. The incorporation of bursts into the analysis process significantly improved the capability to distinguish sounds produced by beetle larvae from background noise when these insects were hidden in stiff, fibrous structures, and likely will be of assistance also in other applications where consistent activity patterns of hidden pests can be identified.