Parameter estimation of the Lotka-Volterra type ecological system involving the prey and the predator is considered. Instead: of switching to a quadratically convergent Gauss-Newton method after reaching the vicinity of a local optimum, the direct search optimization procedure as developed by LUUS and JAAKOLA [1] is used throughout. To obtain rapid rate of convergence of the direct search optimization procedure using random numbers and search region contraction, the LJ optimization procedure is used in multi-pass fashion, where the region size for each variable at the beginning of each pass after the first one is determined from the extent of the variation of the variable during the pass, as suggested by LUUS [2]. Such an approach enables the optimum to be determined very accurately and efficiently. It is found that numerous local optima exist, so that for the given data, a large set of parameters give almost the same accuracy of estimation.