The use of phase contrast optical microscopy for the measurement of airborne fibers using an upper diameter limit of 3 microns was compared with a sampler designed to collect fibers according to the thoracic definition based on the aerodynamic diameter of compact particles. In the absence of commercially available samplers suitable for thoracic sampling of fibers, penetration of several size distributions of fiber through a hypothetical thoracic sampler using the American Conference of Governmental Industrial Hygienists thoracic definition was calculated. Comparison of the results of microscopic aerosol analyses conducted in several facilities using refractory ceramic fibers, with the calculated results, demonstrated that for the most probable fiber size distributions, the physical 3 micron diameter selection criterion was in good agreement with the thoracic definition. The best agreement was seen for small median fiber diameters (less than or equal to 1 micron); significantly lower counts relative to the thoracic definition were seen for large median fiber diameter (8 microns). The author concludes that a thoracic sampler would have several advantages over microscopy; however, until such a sampler becomes commercially available for fibrous aerosols, incorporation of the 3 micron upper diameter limit into the criteria used for fiber counting may be a sufficient surrogate for thoracic sampling.