ABSTRACT. Vertical zonation of hard corals in tropical coral reef Systems is a weil documented and predictable pattern in the organisation of tropical coral communities. However, the mechanisms that underly vertical zonation of hard corals are poorly understood. Two alternative conceptual models for the maintenance of vertical zonation are considered; depth-dependent settlement of larvae, or indiscriminant settlement of larvae with distnbutions determined by differential post-settlement mortality with depth. These alternative models were evaluated by undertaking field translocation experiments using 10 d old juveniles of 3 species with contrasting adult distnbutions: Goniastrea aspera (Verrillrestncted to shallow reef flat habitats; Oxypora lacera (Verrili)-restncted to low light, lower slope habitats; and Platygyra daedalea (Ehs and Solander) -a species with a broad depth range. Laboratoryraised larvae of each species were settled onto terracotta settlement plates and the plates translocated to 3 depths in the field: reef flat (0 m), mid slope (5 m), lower slope (10 m). Survival of the laboratorysettled juveniles was monitored at 3, 6 and 12 mo after settlement. For each of the 3 species, juveniles survived equaliy well at all 3 depths over the 12 mo study penod. The pattern of survivorship of G. aspera and P. daedalea among depths, however, was dependent on plate surface (top, bottom). Given the absence of depth-dependent survival of juveniles from the 2 zone-specific species G. aspera and 0. lacera, we reject the hypothesis that adult distribution patterns of these 2 species is determined by indiscriminant settlement of larvae followed by differential early post-settlement mortality with depth.