In this study, the natural environment of Poecilia reticulata (Guppy) was simulated to estimate an efficient guppy-mosquito larvae density that can be used for routine biological control practice. The fish predatory pattern was evaluated in the presence of alternative preys and their predatory behaviour closely observed for 24 hours. Varying prey densities were exposed to varying densities of P. reticulata. The interaction between P. reticulata and the various prey types were assessed using 2-way ANOVA test and Manly’s selectivity index (S). Prey consumption of P. reticulata increased significantly with predator-prey density and time interval (p<0.05). Individual fish of similar sizes attained satiation at predator-prey density of 1:30 over the period of observation. Despite the preference for Chironomus larvae, P. reticulata never ceased feeding on mosquito larvae. Further analysis with Duncan post hoc tests showed that the four time points of application (6 hour, 12 hour, 18 hour and 24 hour) were the same only at the lowest and highest predator densities of 1 and 10 respectively. There were variations in the peak periods of mosquito larvae consumption by the fish that followed with decline in feeding even with increased mosquito larvae density. The observed predatory behaviour of the fish revealed shoaling behaviour; avoidance of immobile larvae; low foraging when single; competition for prey when paired; total avoidance of mosquito pupae; and more time spent on surface water. It is most likely that the presence of alternative preys will affect the efficiency of P. reticulata as mosquito control agent under field condition. The right estimation of predator-prey density may increase the sensitivity of guppy by facilitating total consumption of larvae within a time span.