Fish received increased attention as a potential source of animal protein and essential nutrients for human diets. Artificial food is required to maintain a high density of fish in farming conditions, than the natural fertility of the water. Fish reared in farming systems have different nutritional requirements than those in the wild. So, well-balanced feeds are necessary nutritionally for pond culture. Growth rate, efficiency of feed conversion and chemical composition of fish have a pronounced effect on the quantity and quality of food consumed. The growth and feed conversion ratio of a fish is remarkable tool to compute the acceptability of artificial feed. Thus, the aim of the present study was to investigate the growth performance, proximate and elemental composition of Catla catla reared under polyculture system by feeding artificial fish diets containing different levels of crude protein. For the present study, four different artificial fish diets with 15% CP (CP-15), 20% CP (CP-20), 25% CP (CP-25) and 30% CP (CP-30) were prepared by using locally available ingredients. Feeding trials were conducted in twelve earthen ponds at three different sites of southern Punjab, Pakistan (Figure I), during July to December, 2014. Catla catla were stocked at the rate of 3942 fish/ hectare (2000 fish/acre), having 30% of studying Major Carp, Catla catla, with Labeo rohita (60%) and Cirrhinus mrigala (10%) in the ponds. Four different artificial diets containing 15%, 20%, 25% and 30% of crude protein (CP-15, CP-20, CP-25 and CP-30, respectively) were offered at the rate of 4% of body weight of fish stock in four different treatments, T1, T2, T3 and T4, respectively. To adjust the feeding rate and to observe growth of fish, sampling was done at an interval of 30 days. The feeding trials lasted for 180 days. Various growth parameters such as percent weight gain, specific growth rate (SGR) and feed conversion ratio (FCR) were calculated by using standard formulae, at the end of feeding trial. Proximate composition of fish samples from each treatment was analyzed. Water content was determined by the oven drying method. Ash content was determined by using a muffle furnace. Fat content was measured by extraction in a mixture of chloroform and methanol. Protein content was estimated by difference from the mass of other main constituents. Elemental concentrations of Cd, Fe, K, Na, Pb and Zn in the fish samples of Catla catla were determined by using Flame Atomic absorption spectrophotometer. Differences among treatment means were determined by Duncan’s Multiple Range Test at a p<0.05 level of significance. Presently reported study suggested that among the different dietary protein levels (15%, 20%, 25% and 30% CP), tested for the study of growth performance of Catla catla under polyculture system with high stocking density (2000 fish/acre), 25% CP showed better FCR, survival and production. Hence, best protein level for optimum growth of Catla catla was observed to be 25% and it was not significantly different from that achieved by 30% (p<0.05) dietary protein level. Hence feed containing 25% CP was found to be satisfactory for fish growth instead of 30% CP. The present study also revealed feasibility of higher stocking density (upto 2000 fish/acre) especially in polyculture system. Estimated slope (b) values for LWRs in all four studied treatments and for overall data were found within the expected range of 2.5 - 3.5. Results for LWRs indicated negative allometric growth pattern in T1, positive allometric in T3 and T4, while isometric growth pattern in T2 for overall data. Mean percentages of moisture, ash, fat and protein contents in different treatments were ranged 74.82-77.05, 3.26-4.51, 5.10-5.18 and 13.52-15.94%, respectively. The dietary protein level 25% and 30% in fish feed would result in quality improvement of the chemical composition of fish. Proximate body composition of Catla catla of treatments having 25% CP and 30% CP were similar and found to be nonsignificantly higher than those fed with 15% and 20% CP. Hence feed containing 25% CP was found to be satisfactory for fish growth instead of 30% CP. Predictive regression equations in the present study demonstrated that body size significantly influenced the proximate body composition of Catla catla. Mean Fe, K, Na and Zn concentrations (μgg-1) in wet weight of Catla catla were found 58.58-95.05, 237.89-262.71, 379.44-396.78 and 14.47-16.55 μgg-1, reared under polyculture system by feeding various levels of fish feed in different treatments. Cadmium and lead were found below detectable limits in the body of Catla catla in all studied treatments. Analyzed metal concentrations were found lower than the permissible limits, proposed by different international organizations. For all studied treatments, the metal concentrations increased in the order: Cd=Pb<Zn<Fe<K<Na in the present work. Elevated trend was noticed among the essential metals and decreased trend among the nonessential metals were observed in the fish. Potassium concentrations in whole body of Catla catla presented differences (p < 0.05), however, iron, sodium and zinc concentrations showed no significance difference among studied treatments. Moreover, size of the fish found significantly correlated with Na and K, while remained constant with Fe and Zn concentration. The present work will help to improve fish production as well as promote aquaculture industry in the country. It will also be helpful in popularization of high stocking fish culture and use of artificial diets in public and private sector within the country. However, further studies on growth performance and chemical composition of fish by feeding various levels of dietary protein levels and inclusion of other fish species and hybrid fish in polyculture farming system may help in enhancing the production.