Green building is a recent design philosophy that requires the consideration of resources depletion and waste emissions during its whole life cycle. Simulation-based optimization can assist designers to achieve a better building design by overcoming the drawbacks of trial-and-error with simulation alone. This dissertation presents the design and implementation of a simulation-based optimization system for the conceptual design of green buildings.  In the optimization model, variables are mostly envelope-related design parameters such as orientation, building shape, wall type, and wall layer. The concept of structured variable is used to describe the hierarchical relationship between variables. Life-cycle cost and life-cycle environmental impact are two major objective functions that respectively evaluate the economical and environmental performance of a building. The impact categories considered in this research include resource depletion, global warming, and acidification. They are unified together with the indicator "expanded cumulative exergy consumption", which is calculated as the sum of the cumulative exergy consumption due to resource inputs, and the abatement exergy consumption due to waste emissions.  The system consists of four components: the input and output, the optimizer, the simulation programs, and the data files. The genetic algorithm is implemented in the optimizer to solve both single- and multi-objective optimization problems. The simulation programs are developed based on the ASHRAE toolkit for building load calculations in order to evaluate objective functions and functional constraints. The system is developed with the object-oriented technology. An object-oriented framework, which is a reusable software architecture represented by a set of classes, is proposed in this research to facilitate the reuse of code and software design. This framework can act as a basis to solve many other simulation-based optimization problems.  A case study is used to demonstrate the application of the system. In this case study, a multi-objective genetic algorithm is employed to optimize a single-story office building in terms of the life-cycle cost and life-cycle environmental impact. The case study resulted in multiple Pareto solutions which can help designers to understand the trade-off relationship between reducing environmental impacts and increasing costs due to green design strategies