In aquaculture, a thorough knowledge of the specific environmental requirements of a species is needed in order to maximize growth rate and survival. There is a paucity of data regarding the fundamental environmental requirements for the ongrowing phase of clownfish juveniles. This prompted the design of three experiments to determine the best feeding regimen, temperature and stocking density that maximize growth and survival of Amphiprion percula. Ration size and feeding frequency are important factors for optimizing fish growth during the juvenile grow-out phase. A factorial growth trial was conducted to determine the effect of feeding frequency and ration size on the growth of juvenile clownfish (Amphiprion percula).Three feeding frequencies (1, 2 and 3 times daily) and six rations (2,4,6,8, 10 and 12 % body weight per day (B W. day-I)) were used to test the growth response over a twelve week period. Non-linear regression analysis on the effect of ration, independent offeeding frequency, on growth resulted in a significant (n = 36; r = 68.7) parabolic model: In y = -0.0302x2 + 0.5159x + -4.4377. Maximum growth corresponded to a ration of8.5% BW.day-l. Survival as a function of ration was significantly lower at2% BW.day-l. Data were further examined with Analysis of Co Variance (ANCOVA) to determine the effect of ration on growth at each feeding frequency. The combination revealed a maximum growth rate when the fish were fed a ration of 10% BW.day-' divided into two equal meals. The required ration per meal to maintain maximum growth was also found to decrease as feeding frequency increased. Abstract The determination of the best temperature for growth is of great importance due to the direct relationship between fish metabolism and temperature. The thermal preferendum of A. percula has already been estimated at 26 ± 0.7°e but it is not known whether this closely approximates the temperature for maximum growth. Sixteen tanks were set to different temperatures ranging between 21.5 and 30.2 °e. Ten juvenile A. percula were placed in each tank and growth was measured fortnightly over the course of the 10 week experiment. Non-linear regression analysis resulted in significant models for fish length (y = -0.0005x2 + 0.00267x - 0.0338; ~ = 56.7, n = 11) and weight (y= -0.00016x2 + 0.0084x - 0.1073; ~ = 61.6, n = 11). These models predict that maximum growth would be at 27.7 and 27.1 °e for length and weight, respectively. Temperature, over the range tested, had no apparent effect on survival. There was no significant difference between the temperatures for maximum growth and the preferred temperature. The effect of stocking densities ranging from 0.2 fish.L-1 to 4.0 fish.L-1 on growth were used in the third experiment. The fish were fed to satiation twice daily and growth was measured fortnightly throughout the 8 week experiment. No effects on growth, survival or coefficient of variation were found within the range of stocking densities tested.