Empirical validation of building simulation results is a complex and time-consuming process. A well-structured and thorough experimental design is therefore a crucial step of the experimental procedure. A full-scale empirical validation study is planned to take place within IEA EBC Annex 71: "Building energy performance assessment based on in situ measurements". The experimental data are currently being gathered in two experiments being conducted at the Fraunhofer IBP test site at Holzkirchen in Germany. This paper describes the methodology followed during the experimental design of the project. Particular focus is on how Building Performance Simulation (BPS) was used to assist the preparation of the actual experiment, to determine suitable test sequences, magnitudes of heat inputs and temperature variations. A combination of both deterministic and probabilistic simulation (using the method of Morris) is employed to replicate the actual experiment and to assess the sensitivity of the model to uncertain input parameters. A number of experimental errors are identified in the experiment, primarily concerning the magnitude of heat inputs. Moreover, the paper includes a discussion on lessons learned from the simulations and on the reliability, reproducibility and limitations of the suggested experimental design procedure.