Optimizing greenhouse design is a complex challenge that involves various design elements, their interactions with crops and the outdoor climate, and large solution spaces. Moreover, evaluating greenhouse performance requires accounting for economic and environmental dimensions, as well as different stakeholders priorities. To address this challenge, this paper made a novel combination of operational research methods with bio-economic modelling. Specifically, a bio-economic model was used to simulate the yield, energy use, and economic and environmental performance of different greenhouse designs. A genetic algorithm was used to explore the large solution space to reduce the computational effort. The overall performance of greenhouse design was evaluated using a directional distance function approach, which incorporated stakeholders’ priorities for economic and environmental performance through the directional vector. The results identified a range of greenhouse designs that were efficient for both investors and policymakers under different price scenarios, for four locations in China. Categorical regression analysis revealed that lighting system, structure, thermal screen, and CO2 dosing are the most influential factors on operating income. Lighting is the primary contributor to greenhouse gas emissions, while incorporating thermal screens effectively reduces emissions.