In southern Taiwan, brooded larvae of Pocillopora damicornis and Seriatopora caliendrum are released year-round in synchrony with new moons, and each larval release occurs over multiple days. Using P. damicornis and S. caliendrum as a model system, we describe within-brood variation in larval phenotypes and test for release-day effects that influence larval performance in the pelagic phase. Research was conducted in 2010 using larvae from corals collected in June and July from Nanwan Bay (21°56. 179′N, 120°44. 85′E). In June, larval phenotypes of both species were characterized immediately following release, and their competency to settle assessed. In July, larvae of P. damicornis were collected on 3 days over the peak release period and incubated for 7-11 days at 28. 0 °C and 320 μmol quanta m-2 s-1; their phenotypes and settlement competency were measured every 2 days. P. damicornis larvae released close to peak release were 1. 6 times larger in size, contained twice the number of Symbiodinium larva-1, and were 44 % more likely to settle in the first 24 h than larvae released early in the brood. In addition, peak-release larvae respired at a lower rate than larvae released late in the brood. Similarly, S. caliendrum larvae released close to peak release were 1. 4 times larger in size and were 33 % more likely to settle in the first 5 h than larvae released early in the brood. In July, P. damicornis larvae differed between early (2 days prior to peak), peak, and late (2 days after peak) release. Protein content of early-release larvae was lower than peak- and late-release larvae, and this difference persisted throughout the development. Further, release day affected the way larval respiration varied throughout development. By showing that brooded coral larvae differ between release days and display maternal effects influencing performance in the swimming phase, our results suggest that pocilloporid corals utilize bet-hedging to increase reproductive success.