: A life-size model human head and shoulders was exposed in a wind tunnel to winds between 0.75 and 2.75 ms/s and particles up to 30 micron aerodynamic diameter, and the weight of material entering the nose and mouth at various breathing rates was compared with that caught by isokinetic samplers. When aspirated at a constant rate, the head behaved like a blunt sampler, showing a ten-fold variation in entry efficiency for different windspeeds and directions, and a two-fold variation even at 5 micron. Detailed facial structure had little effect. The efficiencies for various breathing patterns and minute volumes were calculated, and when averages were taken for all wind directions, to simulate a worker randomly oriented to the wind, the efficiencies lay within a fairly small range, and were all substantially less than 100%. "Total" dust standards and samplers should take account of these efficiency curves.