Multiple-site-type catalytic systems produce ethylene/1-olefin copolymers with broad molecular weight distribution (MWD) and chemical composition distribution (CCD) because each active site type produces chains with distinct chain microstructures. Simultaneous deconvolution of the MWD and CCD can be used to identify the number of active site types and chain microstructures produced on each active site type by performing parameter estimation to minimize the sum of the squares of differences between experimental and model data. However, conventional optimization algorithms often rely on the adequate initial guess. In this work, a genetic algorithm, which is a stochastic optimization search heuristic that mimics the natural evolution process, was implemented during the simultaneous deconvolution. The proposed approach was validated with simulated data of model ethylene/1-butene copolymers produced using a system with three active site types.