This study was designed to investigate the effects of starvation and re-feeding cycles on the growth performance and body chemical composition of Oncorhynchus mykiss juveniles. A total of 360 juveniles with initial mean weights (IW) of 8.46 AE 0.07 g (n = 360) were stocked into 400-L tanks in triplicate for each group, with 30 juveniles per tanks. The control group received regular feed, as is the common practice. The three other groups were periodically starved: 1 day starvation followed by 6 days re-feeding (S1), 2 days starvation followed by 5 days re-feeding (S2) and 3 days starvation followed by 4 days re-feeding (S3). The experiment lasted for 10 weeks, over the course of which the water flow rate was 4 L min À1 and the water quality parameters determined as: temperature 14.4 AE 1.1°C, oxygen 8.2 AE 0.4 mg L À1 and pH 7.5 AE 0.2. At the end of the study, S1 had the best growth performance (final weight, specific growth rate, average daily growth) of all test groups (P < 0.05). The lowest daily feed intake (DFI) and growth performance parameters were observed in S3 (P < 0.05), while protein efficiency ratio (PER), net protein utilization (NPU) and lipid efficiency ratio (LER) were higher in the S3 fish group than in the other groups (P < 0.05). Whole body protein and lipid contents were highest in S1 fish. The hepatosomatic index (HSI) and viscerosomatic index (VSI) were significantly different among groups (P < 0.05). Feed conversion ratio (FCR) was significantly lower in starvation groups S1, S2 and S3 than in the control (P < 0.05). Compensation coefficient (CC) values were higher than 1 in all starvation groups. The concluding indicate that rainbow trout exposed to 1 and 2 days of starvation in week cycles could achieve over compensation compared to the control. Additionally, partial growth compensation and improved feed utilization could be achieved in a starvation group within 3 days in a week, by beginning with the juvenile size over a 10-week experimental period.