Predation was examined using experimentally deployed shells along two transects representing shallow shelf (15 to 30 m), outer shelf (70 m) and bathyal (100 m to 267 m) habitats in the Bahamas. Lethal breakage on experimentally tethered shells was not restricted to the shallow-shelf (≤ 30 m), but could occur to outer shelf (70 m) and upper slope (≥ 88 m) depths. However, significantly more shells were preyed on at shallow-shelf depths (≤30 m) than at deeper depths (≥ 70 m) for both transects. Predation was not restricted to shelf sites (≤ 70 m), and predator-induced damage, such as peeled shells or last whorl remnants, could be encountered to a depth of 195 m. However, significantly more predation occurred at shelf rather than slope (≥ 88 m) depths for both transects. Although the experimental shells were exotic species, they were preyed upon at differing frequencies depending upon transect location and depth. It appears that morphology alone may not account for the differences in predation. Long-term deployment (six years) of shells enhanced the likelihood of predation, but short term deployment (one or two years) did not. Molluscivorous predators in the Caribbean are highly diverse and are not restricted to the shallow shelf, but occur in outer shelf to slope settings. Therefore, evolutionary-ecological conclusions based on predation with depth should not be tied to a single site, to a few depths, or to a single predator, as habitat heterogeneity and patchiness in predator and prey distributions may vary.