Predator-prey interactions have a major effect on species abundance and diversity and aggregation is a well-known anti-predator behavior. For immobile prey, the effectiveness of aggregation depends on two conditions: (a) the inability of the predator to consume all prey in a group and (b) detection of a single large group not being proportionally easier than that of several small groups. While the benefits of grouping to avoid visually hunting predators are well understood, the potential costs and benefits of aggregation when visual cues are not available are not well understood. We carried out foraging (predation) experiments using a fish predator and (dead) chironomid larvae as prey in both laboratory and field settings. In the laboratory, a reduction in visual cue availability (in turbid water) led to a delay in the location of aggregated prey compared to when visual cues were available, but aggregated prey suffered high mortality once discovered, leading to better survival of dispersed prey in the longer term (this was likely due to their inability to take evasive action and due to prey groups being small). In the field (where prey were placed in feeding stations that allowed transmission of olfactory but not visual cues), aggregated (large groups) and semi-dispersed prey survived for longer than dispersed prey – including long term survival. Together, our results indicate that like in systems where predators hunt using vision, aggregation is an effective anti-predator behavior for prey avoiding olfactory predators.