Management of commercial sea cucumber stocks relies, in part, on estimates of population densities which, in-turn, depend on knowledge of habitat preferences, and of the influence of biological cues on sheltering and/or aggregation behaviour. Here, we document a diurnal shift in the sheltering behaviour of the Pacific and eastern Indian Ocean black teatfish, Holothuria whitmaei, and discuss the implication of this behaviour for surface-based population density surveys. Diurnal studies of 30 black teat fish on Ningaloo Reef, Western Australia, found that the proportion of animals sheltered (and therefore hidden when viewed from directly above) was significantly greater in the morning (3–23%; AM_0830-1230 h) relative to the afternoon (0–6%; PM_1230-1730 h). As with sheltering behaviour, the straight-line distance between individual sea cucumber and the nearest shelter also showed marked diurnal variation, with animals observed at greater distances from shelter between 1230 and 1730 h (PM_4–22 cm; AM_1–7 cm). Based on these results, we suggest that surfaced-based census techniques (e.g. manta tows) may underestimate population densities if conducted during times of reduced activity and increased incidence of sheltering behaviour. Appropriate calibration factors for day-time black teatfish surveys are proposed.