Industrial hygiene particulate samples are often collected under anisokinetic sampling conditions and in crosswinds. Experiments were conducted to quantitate errors associated with sampling under these nonideal conditions. Three types of field sampling cassettes were tested to determine particle sampling efficiencies for 0, 2, and 5 m/s winds at incidence angles of 0, 90, and 180/sup 0/. Tests were conducted in a chamber in which the cassettes or filter holders were moved to generate a relative wind. Eight filter cassettes mounted symmetrically on a horizontal arm which rotated about a vertical shaft at 250 rpm were positioned so that four cassettes moved at 2 m/s and four at 5 m/s. In addition to two each of the test cassettes, two in-line filter cassettes equipped with isokinetic probes were mounted on the arm. Mass, or number of particles, collected by test cassettes and isokinetic samplers was compared to that collected by stationary filters placed near the rotating arm. Results from tests in the rotating arm chamber were checked against similar tests made in a turbulent flow wind tunnel. Sampling efficiencies were compared to theoretical values calculated from Belyaev and Levin's theory, and Davies' modification for calm air sampling. Satisfactory agreement was obtainedmore » for 5-..mu..m D/sub ae/ particles; however, considerable discrepancy was noted with larger particles. The results indicate that corrections must be made to determine concentration when particles collected under nonideal sampling conditions are larger than approx. 5-..mu..m D/sub ae/. « less