Flow apportionment of a 8-parameter flow manifold is dealt in this paper. We deal with a one-by-four manifold. We have to optimally locate the guide plates. The objective of the work is to minimize the standard deviation of the actual flow rates from the set points by controlling 8 parameters four of which are lengths and four of which are angles of deflection. We take the input as the 50 data points generated by the process of CFD. Using the combination of ordinary least squares and genetic algorithms we develop the minimization algorithm for the objective function. We have successfully evaluated the objective and 8 parameters and found that the algorithm employed in this work yields a solution that is lower than the lowest of the 50 data points. Therefore we concluded that our method works successfully without having the need to resort further to CFD computations. © 2017, North Atlantic University Union. All rights reserved.