This paper describes the suitability of a closed recirculation system to study the development of fish larvae in a strictly controlled environment, where only feeding was varied (fed, starved, delayed and late fast treatments). The system served both as an incubator and hatchery. The time variation of physical and chemical parameters together with survival and growth of reared sea bass larvae (Dicentrarchus labrax) were studied over the first month of life. The recirculation design allowed for the maintenance of levels of ammonia, nitrite and nitrate below those cited as responsible for mortality or decreased feeding ability in other marine fish larvae. Almost no larval mortality occurred in the fed larvae from day 9 after hatching. The starved group of larvae showed a sharp decline in survival after 16 days of food deprivation. Larvae for which feeding was delayed until day 13 ceased dying 4 days after food was supplied. Fast feeding on days 24 and 25 had no effect on larval survival. Growth in length was similar during the first 2 weeks of larval life regardless of feeding treatment. Two days of late fast had no effect on larval growth. Growth patterns of fed larvae in this study were similar to those reported for larvae reared in flow through systems. We believe that survival and growth of the reared larvae was a direct function of diet, and that the type of rearing system did not adversely affect these parameters. The rearing design and the results obtained suggest that future nutritional studies of field-collected larvae will benefit from this kind of rearing experiment.