A series of laboratory and field experiments was carried out to determine the effects of gamete dilution on fertilization rates in three species of reef coral. Gametes remained viable for 2 h after spawning, but one species exhibited signs of reduced fertility 3-4 h after spawning. Sperm dilution trials carried out in the laboratory indicated that fertilization reaches a maximum at sperm concentrations of 10(5)-10(6) per ml, with reduced fertilization occurring at both higher and lower concentrations. Estimates of ''fertilization potential'' in the field were obtained by exposing eggs to water samples taken from the field at various times and locations following episodes of coral spawning. This sampling program indicated that on nights when only small numbers of coral spawned (minor spawning), the fertilization potential was much lower than on major spawning nights. On major spawning nights, fertilization potential was consistently high just after spawning, but became spatially variable thereafter. The percentage of fertilization in field-collected samples of eggs and embryos just after spawning was also higher during major spawning nights than during minor spawning nights. These measurements indicate that gamete dilution can play an important role in limiting the fertilization of coral eggs in the field during natural spawnings. It follows, therefore that corals are under considerable selective pressure to spawn synchronously in order to generate high gamete concentrations in the water column and thus to maximize the probability of successful fertilization. In addition to spawning in synchrony, corals also minimize the effects of gamete dilution by spawning buoyant gamete bundles that accumulate at the sea surface, and by spawning during periods of low water motion.