Abstract The feeding inefficiencies associated with intensively cultured prawn systems have a significant financial cost and environmental impact. Initial trials of a commercial system using sound to manage feeding within cultured systems have achieved promising results with an impressive food conversion (food weight/biomass) ratio of 1.42. Whilst these results demonstrate the potential benefit of employing passive acoustics for feed management, the underlying technologies are not well understood by industry or the research community. Consequently, a sound based study of feeding tiger prawns is conducted to investigate the key challenges associated with passive acoustic approaches; sound detection and feed demand estimation. The study finds that tiger prawns produce impulsive sound signatures during feeding that can be used as a proxy of feeding activity. Spectral features of the feeding signatures can be used to detect feeding activity within acoustically complex farm ponds (feeding signal to noise ratios less than −40dB), given they are largely disjoint from the background noise spectrum (aerators) across all farm recordings. One of the potential challenges identified with sound based feed detection is that other sources of event driven interference arise (i.e. rain and faulty aerators), which can be misclassified as feeding. Whilst our investigation indicates that sound based detection of feed events are plausible, it is unclear how accurate it is to estimate the quantity of consumed pellets from feeding sound. Our study provides evidence to support its feasibility, given the temporal evolution of feeding sounds and pellet consumption were linearly related across tank and pond based feeds, respectively (R2=0.95 and R2=0.96).