Study area and period: The experiment was conducted at the Noakhali Integrated Agro Industries Limited (NIAIL) located in between 22°38´and 22°59´ north latitudes and in between 90°54´ and 91°15´ east longitudes in the Sadar upzilla of Noakhali district, Bangladesh for a period of six months from May to November. Species selection for mono and polyculture: Three fish species such as Nile tilapia, Oreochromis niloticus, Climbing perch, Anabas testudineus, Asian stinging catfish, Heteropneustes fossils were selected for polyculture with M. gulio. Experimental design: Nine rectangular earthen ponds of 1.0 decimal, having an average depth of 1.5 m were used for the experiment. Three treatments (T1, T2 and T3) each with 3 replications were designed according to completely randomized design (CRD). Collection of fish, stocking and feeding management: Fry of M. gulio was collected from nursery ponds of NIAIL, Noakhali. Prior to release in experimental ponds, fry were acclimatized by a short 5 to 10 seconds bath and prophylactic treatment with (5% KMnO4) solution. Fry of M. gulio, O. niloticus, H. fossilis, and A. testudineus were counted and stocked in triplicates ponds of different treatment (Table I). The fishes were hand-fed to apparent satiation four times per day (at 8.00, 11.00, 14.00 and 17.00 h) during the first month, thrice daily (at 9.00, 13.00 and 17.00 h) in second months and then twice a day (at 9.00 and 17.00 h) from third month to the end. During feeding time, a close observation was made and a record of supplied feed was kept for determining the apparent feed conversion ratio (AFCR) and apparent protein efficiency ratio (APER). The proximate composition of the feed, used in the current experiment was analyzed in triplicate according to standard procedures by AOAC (2000). Sampling of fish and monitoring water quality: A fortnightly sampling of fish was made by using a cast net and weight of fish measured by using a digital balance (OHAUS, CT 1200-S, USA) and length by a centimetre scale. Water quality parameters viz. temperature, dissolved oxygen (DO), water pH, soil pH, salinity, transparency, nitrate (NO2), nitrite (NO3), phosphate (PO4), total ammonia (NH3) and alkalinity were estimated and recorded on spot throughout the experimental period by using standard procedures and methods. The water temperature (°C) was measured by using a standard mercury thermometer, DO (mg/L) by DO meter (YSL, Model 58, USA), water pH by digital pH meter (Elico-Li-120), soil pH by soil pH and moisture tester (KS-05, TLEAD, China), salinity (ppt) by a refractometer (ATAGO, Japan), transparency (cm) by Secchi disc. Nitrate (mg/L), nitrite (mg/L), phosphate (mg/L), ammonia (mg/L) and total alkalinity were measured by using universal pocket meters. General conditions of the ponds were monitored regularly and the water quality parameters were measured between 10.00 to 13.00 h during the culture period. Calculation of growth parameters Mean final weight gain (g)= {Mean final fish weight (g) - Mean initial fish weight (g)} % weight gain= [{(Mean final fish weight (g) - Mean initial fish weight (g)}/ Mean initial fish weight (g)] x100. Specific growth rate (SGR %/day)= [{(Loge W2 - Loge W1)/ (T2-T1)} ×100], Where, W1 is the initial live body weight (g) at time T1 and W2 is the final live body weight (g) at time T2 (day) Apparent feed conversion ratio (AFCR)= Total dry feed fed/Total live weight gain . Apparent protein efficiency ratio (APER)=Live weight gain/Dry weight of crude protein fed. Production (kg/ha) = [{Final number of fish harvested x individual weight of fish (g)}/1000] x 247.1. Statistical analysis: One-way ANOVA (Duncan 1993) was applied by performing SPSS 17 for Windows (Chicago, USA). For all statistic tests,a=5% was chosen as the significance criterion.