Broadcast spawning of gametes with planktonic development of larvae is the most common reproductive mode in tropical corals, and is generally thought to optimize the dispersal potential of larvae. To this end, many previous studies of coral larval dispersal have focused on the maximum time larvae can remain competent to settle and consequently how far they might disperse. However, dispersal ability of broadcast-spawned coral larvae will be linked, at least in part, to the minimum time to settlement competency as well as the length of the planktonic period - although estimates of minimum time to competency remain largely anecdotal, with few rigorous studies of the pre-competent period. To determine the minimum time to larval settlement in two species of broadcast-spawning coral (Platygyra daedalea and Goniastrea favulus), we monitored larval settlement rates in aquaria every 6 h from the time larvae commenced swimming (i.e. were ciliated, fully developed larvae) for a period of approximately 10 days. For P. daedalea, peak settlement occurred between 60 and 66 h following fertilization (2.5 and 2.75 days), which is markedly earlier than the 4- to 6-day time period commonly cited as the minimum time before broadcast-spawned coral larvae are competent to settle. Surprisingly, it was also clear from our experimental results that settlement in P. daedalea occurred as a distinct pulse during the 60- to 66-h period, rather than continuously throughout the study period. G. favulus larvae also appear to be able to settle quickly (from 54 h following fertilization). We argue, on the basis of these short competency times and apparently rapid settlement, that dispersal in broadcast-spawning coral larvae may not be as great as has previously been assumed.