A study was planned in order to improve the knowledge on heat, air and moisture transfers within light-weight buildings envelopes. In this aim, an experimental cell exposed to real climate, consisting of a small wooden frame structure and panel walls is being built. Its dimensions correspond to a common living-room (25m², 2,50 m high); the room can be heated, ventilated, and vapour can be produced. Instrumentation comprises the measurements of temperature and humidity, in the room and within the walls, as well as energy consumption and outdoor conditions. The aim of this large scale experiment is to evaluate the influence of different parameters, such as the thickness of the insulation, the thermal inertia, the vapour-barrier, etc. on the indoor environment and building structure. The whole study consists of 3 phases: § Design of the experiment using numerical simulation § Large scale experimentation and model validation/improvements § Parametric study using numerical model The numerical model used is based on the HAM-Tools, developed by Chalmers University (Sweden) and the Technical University of Denmark. This tool is a component library for use in the Matlab-Simulink environment, allowing modelling the whole building with one node per thermal zone, and a one dimensional model of the walls. It allows the user to define a ventilation rate, internal loads, and/or heating/cooling devices, and gives the results in terms of energy consumption, air temperature and relative humidity in the zone, temperature, relative humidity and moisture content within the walls. This work consisted first in eliminating some a priori irrelevant (or not feasible) sequences of experiments among the large set of all possibilities. The few remaining ones were then tested by performing the different simulations involved in each sequence. The last task was to choose the sequence that showed the best difference(s) between its different steps, meaning that the effect of each tested parameter was more visible in this order of experiments. This phase is very important in order to optimise experimental investigations. Indeed, switching between steps requires serious modifications of the experimental cell, such as adding or removing additional insulation layer and/or vapour or air barrier, etc. Also the duration of each step is rather long, several weeks will be needed. It is then necessary to make sure that each tested step, and the sequence of steps, will be relevant, and no experimental time and effort will be wasted. The next phase is to perform the chosen sequence of experiments; the climate measured data will be used as an input for the numerical model, then the simulation results will be compared to the measures, and the model will be adjusted. Then, we will use the numerical model to perform the parametric study itself, in the same conditions for the different configurations. The conclusions should give keys to improve the performance of wooden frame constructions. The paper will focus on the first step; the results and conclusions will be presented, after the presentation of the methodology.