ABSTRACT: We introduce a novel optimization method based on the evolutionary algorithm Particle Swarm Optimization (PSO) to enhance the electromagnetic performance of quadrature hybrid designs. Optimization and simulations batches were conducted using a fully tuned and validated version of the algorithm, for the design of quadrature hybrids intended to operate ALMA (Atacama Large Millimeter Array) Band 2 (67-90 GHz), Band 3 (84-116 GHz), and Band 2+3 (67-116 GHz). Thus, we present quadrature hybrid designs which are optimized to operate in ALMA Band 3 (84-116 GHz) with respect to their operational requirements for the scattering parameters and amplitude imbalance. Furthermore, the resulting designs take into account machining constraints related to cost and feasibility requirements. Finally, this work provides a method that can be easily extended to optimize other microwave devices and waveguides for radio astronomy applications, with the benefit of speeding up the design process as well as reducing the computational costs.