This paper proposes an improved stochastic process model for the selection of categorical variables, such as airfoils and materials, in rotorcraft design. This process leverages trends in typical rotorcraft conceptual design objectives that are similar across different categories. Also, this paper extends the use of efficient global optimization (EGO) algorithms, which intelligently search design spaces, to the previously proposed stochastic process thereby enabling the use of more computationally intensive tools earlier in the design process. To optimize the EGO infill criterion, a genetic algorithm is developed that is capable of searching domains with categorical variables. The proposed stochastic process model is successfully tested against traditional independent surrogates when approximating the engine shaft horsepower of the UH-60A with a choice of airfoils. Finally, a test to optimize the UH-60A engine shaft horsepower at two flight conditions demonstrates that the proposed extension of the EGO algorithm is more efficient at finding the Pareto fronts than the current state-of-the-art.