To develop a feeding strategy for the Australian freshwater ¢sh silver perch (Bidyanus bidyanus Mitchell), a series of eight experiments was done in 1m 3 cages in an aerated, earthen pond to determine the e¡ects of feeding rate (% body weight) and feeding frequency (no. of feeds day À1 ) on the growth and feed conversion ratio (FCR) of ¢ngerlings and larger ¢sh under ambient water temperatures over the range 13.83 0.6 1C. Fish were fed extruded pellets of a silver perch diet containing 34% digestible protein and 14 MJ kg À1 digestible energy. Commercial silver perch farmers were consulted about feeding practices for large ¢sh (4500 g) and at water temperatures below 12 1C, and winter feeding practices for other warmwater species were used to complete the strategy. In the feeding experiments, growth and FCR increased with increasing feeding rates to a level above which only FCR increased. Optimal feeding rates and frequencies were those which resulted in maximal growth, while minimizing e¡ort (feeding frequency) and FCR. The highest feeding frequency required for maximal growth, including that of small ¢ngerlings was twice (2 Â ) daily, and the optimal feeding rates varied with water temperature and size of ¢sh. The optimal daily regimes were: small ¢ngerlings (initial mean weight, 2.0 g) 7.5% 2 Â at a mean temperature of 23.3 1C; ¢ngerlings (14.9^27.7 g) 7.5% 2 Â at 27.1 1C, 5.0% 2 Â at 23.7 1C and 2.0% 1 Â at 16.8 1C; and large silver perch (162.5^510.6 g) 0.5% 1 Â daily or 1.0% on alternate days at 15.6 1C, 1.0% 1 Â at 17.3 1C, 3.0% 2 Â at 24.1 1C and 2.0% 2 Â at 27.9 1C. It is suggested that regimes of 0.5% 1 Â daily for ¢ngerlings (o50 g) and 0.5% 1 Â on alternate days for larger ¢sh are used at temperatures of 9^12 1C, and 0.5% 3 days week À1 and 0.5% 1 dayweek À1 for ¢ngerlings and larger ¢sh, respectively, at 6^9 1C. Feed inputs should not exceed 150 kg ha À1 day À1 in ponds less than 0.3 ha and 100 kg ha À1 day À1 in larger ponds. Our research has established a feeding strategy for silver perch based on restricted rations. Ã Overall mean; there were no signi¢cant di¡erences between means at stocking within each experiment.wRange of mean weights across all treatments and replicates. Aquaculture Research, 2005, 36, 1429^1441 Silver perch feeding strategy S J Rowland et al.