For more than two decades evolutionary algorithms were successfully used in de-novo design of small molecules or in lead structure optimization. Their natural ability to combine the best of two promising solutions into a new one makes them reliable means to explore the vast chemical space. However, as almost all computer-aided drug design techniques do, their solutions often lack synthesizability. Therefore, I developed a new algorithm using novel crossover and mutation operators to efficiently explore chemical make-on-demand libraries to guarantee easy access and in-vitro testing for all solutions.