Listening for engine sounds in an underwater environment is modeled using spatialized infrapitch signals and pink noise maskers. Detection thresholds for such signals were measured for various spatial configurations of maskers. All stimuli were delivered over headphones, and rendered using individually measured Head-Related Transfer Functions. In single-masker environments, no significant effect of signal bearing in azimuth or elevation on detection threshold was observed. In multi-masker environments, the largest detection threshold increase was observed when increasing the number of maskers from 1 to 2. Thereafter, a linear increase in threshold detection (in dB) for every doubling of maskers, up to 32 maskers was observed. Randomizing the spatial presentation of the signal and maskers for each trial did not affect thresholds. Results suggest that prior knowledge of the spatial configuration of the signal and maskers does not appear to be necessary in performing such detection tasks.