The performance of the 10 mm Dorr-Oliver cyclone is known to suffer from bias associated with the dependence of its aspiration efficiency on the inlet orientation. A new multi-inlet cyclone sampler was developed to minimize such orientation bias. The sampling how rate through the new sampler was optimized to match its performance with the newly established ISO-ACGIH-CEN respirable dust sampling convention. Sampling efficiency curves for the multi-inlet cyclone sampler are presented and the cyclone was found to be in good agreement with the ISO-ACGIH-CEN curve in both wind tunnel (at a free-stream velocity of 1.52 m s(-1)) and calm-air chamber tests. At an optimized sampling how rate of 2.5 l min(-1), the sampler has a bias in the mass sampled as low as 0.05%. The new sampler was also designed to be compatible with the existing filter cassette, sampling pumps and the cyclone holder. The samplers were challenged in a wind tunnel using polydispersed coal dust generated in a custom-built fluidized bed. An Aerodynamic Particle Sizer (APS) was used to measure the particle concentration and size distributions collected by the isokinetic sampler and the cyclone sampler. The test matrix consisted of a range of sampling flow rates, orientations from the free stream and free-stream velocities representative of workplace conditions.