A sampling device for the determination of respirable silica (14808607) dust using a 10 millimeter (mm) nylon cyclone and a 37mm cassette loaded with PVC type filter was evaluated. Using a large inorganic sampling assembly, samples of mine dust containing 90 percent quartz were run for 2 to 5 hours. The aerosol concentrations were 2.4 to 3.5 milligrams per cubic meter. Each of the nine samplers used was equipped with a Dorr-Oliver 10mm nylon cyclone in an MSA holder assembly, a two piece 37mm monitor cassette, and a 37mm diameter, 5mm pore size PVC filter. The effect of air leaks in four positions on the performance of the cyclone was tested using five control cyclones and four leaky cyclones. The four positions were: between the vortex finder and the cyclone body; in the lower adapter o-rings; in the upper adapter o-rings; and in the cassette/frame seal. The data was reduced using an analysis of variance that tested the effects of run, leak location, and sampler. Over a deposition range of 826 to 1800 micrograms per sample, the precision was 2.18 percent. In a run with no lower adapter o-rings, leakage was significant. When the lower adapter o-rings were replaced loosely, leakage was still significant. When the upper adapter o-rings were very loose, the leak was also significant. The authors conclude that the 10mm nylon cyclone sampling unit is capable of adequate precision and is rugged with respect to minor air leaks. The authors recommend that the operator check the unit for completeness initially and check for audible leaks during operation.