Measurement of airborne dust concentrations usually involves drawing a sample of the dust-laden air into the measuring instrument through an inlet. Even if the surrounding air is calm, theoretical calculations predict that large particles may not be sampled accurately due to the combined effects of gravity and inertia on the particles near the sampling inlet. Tests were conducted to determine the conditions of particle size, inlet radius, and flow rare necessary for accurate dust sampling. A coal-dust aerosol was sampled simultaneously through inlets of different diameters at the same volume flow-rate and collected on filters. The dust was removed from the filters and the particles were counted and sized with a Coulter counter. Results showed that published criteria for inlet conditions for correct sampling are overly restrictive and that respirable-size particles are sampled correctly in the normal range or operation of most dust sampling instruments.