The critical size hypothesis states that pelagic larvae (vertebrate or invertebrate) must reach a critical size to begin metamorphosis and settle to their suitable reef habitat; however, there have been few critical tests of its broader validity. If there were a narrower range of ages at settlement relative to size, then the data would conform to a critical age hypothesis. It would also be predicted that size or age would be truncated respectively at the lower limit for each hypothesis. These hypotheses were tested for 2 species of coral reef fish from the Great Barrier Reef. Variation in age at settlement forPomacentrus coelestiswas greater than that of size at settlement for fish collected in waters of different temperatures, and the size distribution was truncated at a size of ~9.3 mm; accordingly, we accepted the critical size at settlement hypothesis for this species. In contrast,Scolopsis bilineatusmet one criterion to reject the critical size hypothesis, based on variation, but truncation in size was found. Variation in age aligned with a critical age at settlement hypothesis. There was a variable relationship between age at settlement and size for both species. Strong evidence is provided that growing fast in the plankton and settling quickly may be advantageous for fish. For both species, settlers with a short planktonic larval duration, and sometimes small size at settlement, had grown faster in the plankton and may have experienced better conditions. Differences in pre-settlement growth are likely to affect the maximum size at which fish can settle and post-settlement survivorship.