This paper summarizes part of the work held at ENTPE within the framework of the IEA Annex 35 project. The aim of the work is to develop and experimentally evaluate control strategies for hybrid ventilation systems to provide comfortable indoor environment and good air quality by combining the best aspect of natural and mechanical ventilation. To reach this target and asses the physical model elaborated in this project, a test cell "HYBCELL" has been designed within the laboratory. In order to represent a large office or a meeting room, HYBCELL has been created by totally retrofitting an old local within a large hall whose temperature can be controlled to create an artificial climate around the test cell. The front of the cell is a slopped wall (70) which communicates with the outdoor climate through six sash windows. The test cell is 5.1m long by 3.5m wide and it is 2.9m high. Except the floor, the false ceiling and all walls are made with office building materials. The hall is equipped with a HVAC system creating an artificial climate around the test cell. The temperature can be controlled between 5 and 20C in winter and 15 up to 35C in summer. In order to air-condition and ventilate the test cell, we consider link it to the hall HVAC network and equip it with a blower. To simulate the occupancy, CO2 and sensible heat will be injected in the test cell. Various sensors (Temperature, Relative humidity,CO2 concentration, COV, Pressure) have been installed in the large hall, in the test cell and outdoor. In addition, wind velocity, wind direction and solar radiation can be collected from the meteo station located near the cell. The data acquisition system is based on a DAQ card and the whole experimental device is connected to a Personal Computer. The experimental device will be used to adjust the physical model, characterize the openings (determination the pressure coefficients (CP) and discharge coefficients (Cd)) and carry out several control strategies for hybrid ventilation. On-off and PID Controllers based on temperature, CO2, and PMV control will be the first approach to the problem. Then elaborated controllers based on muticreteria approach will be tested and optimised. This work is financially supported by ADEME (L'Agence De l'Environnement et de la Matrise de l'Energie) and EDF (Electricit De France).