Coupled three-dimensional finite element analysis (3D-FEA) and computational fluid dynamics (CFD) software packages used to simulate heat transfer and airflow can be valuable tools for predicting and comparing the thermal performance of building enclosure components, assemblies, and systems when validated and used consistently. Today, the most commonly used tools are two-dimensional FEA (2D-FEA) programs that have been developed as part of efforts to establish uniform methods of evaluation to compare conduction-dominated heat transfer through building envelope elements, primarily fenestration systems. Increasingly, these tools are being used to predict in-service surface temperatures of building enclosure assemblies, particularly for the risk of interior condensation. However, these commonly used 2D-FEA tools have substantial limitations. They typically do not explicitly model radiation between objects or convective fluid flows, each of which can significantly affect heat transfer at element surfaces. They also cannot account for conditions that vary in three dimensions. As a result of these limitations, the computational tools most commonly used today cannot accurately predict surface temperatures that are significantly affected by convective currents, surface radiation, or three-dimensional variations in relevant factors. Coupled 3D finite element analysis and computational fluid dynamics (3D-FEA-CFD) software packages that can explicitly model convection, radiation and three dimensions are now commercially available, and more advanced analyses can be performed. Because many of these multiphysics packages were initially developed for mechanical applications, relatively little work has been done to evaluate their use in the study and design of building enclosure systems. The objective of this study is to develop and evaluate a practical methodology for using a 3D-FEA-CFD package to predict nighttime in-service building enclosure surface temperatures. The evaluation includes instrumentation and data collection from a test space, 3D-FEA-CFD simulation of the instrumented assemblies, assessment of basic simulation accuracy, comparison to 2D analysis results, and an analysis of sensitivity to selected simulation variables.