A new experimental wind tunnel facility for aerosol sampling investigations has been built and its performance evaluated. Subsequently, an experimental methodology using a polydisperse test aerosol of glass beads to measure entry, transmission and overall sampling efficiencies has been developed and tested. The new facility is composed of a horizontal cylindrical pipe of 5 m long and 30 cm in diameter. The measurement zone is located just at the exit, allowing to take benefit of the whole cross-sectional area inside a stabilised aerosol flow. The working air velocity range is 0.5-4.5 ms(-1). Air Velocity and turbulence profiles are uniform within 10%. Turbulence in the working section is controlled with a square mesh grid. The test aerosol is generated by a fluidized-bed generator and dispersed into the clean air flow upstream of the horizontal part. Generated particles are within a size interval extending from a few mu m to about 80 mu m in aerodynamic diameter. Tests of time and space stability of the test aerosol in the working section were carried out. They have shown a reasonably uniform spatial distribution and time stability considering the size range of generated particles. The experimental method allows to obtain, simultaneously with the same technique entry, transmission, and overall sampling efficiencies of samplers from several mu m up to 70 mu m in particle aerodynamic diameter with a good accuracy. It is based on the measurement of the distribution of particle number concentration vs particle aerodynamic diameter of deposited and sampled aerosols in a reference probe and in the test sampler. To evaluate both the new wind tunnel facility and the methodology, measurements of the different efficiencies were achieved using a cylindrical sharp-edged thin-walled probe as a test sampler. This evaluation was performed in three steps. At first, the reproducibility of transmission efficiency measurements of the probe working in isokinetic conditions was determined. It appears fairly good between 10 and 70 mu m in particle aerodynamic diameter. Then, the methodology was applied to the assessment of the aspiration efficiency of a probe working in subisokinetic conditions. Finally, a consistency test of the data was proposed and applied to our data; it consists in comparing the mass fractions of collected samples (deposited on the internal sampler walls, collected onto filters) calculated from the efficiency data and the distributions of particle concentrations, with those which are directly recovered after each experiment and weighed. This test yields an indicator of the quality of the whole efficiency data set.