In this work, three types of needlefelt lters, made of Polyester (PE), Ryton Sulfar (RS), and Polyaramid (PA), were tested to investigate the aerosol loading characteristics of fabric lters when challenged with micrometer-sized monodisperse potassium sodium tartrate (PST) particles. A brous lter with packing density of 9%, thickness of 0.38 mm, and ber diameter of 5.1 µ m was included for comparison. A vibrating ori ce monodisperse aerosol generator was used to produce three different sizes (5, 10, and 20 µ m) of PST particles for aerosol loading experiment. An ultrasonic atomizing nozzle and a TSI constant output nebulizer were used to generate polydisperse PST particles for the aerosol penetration test. The aerosol penetration of submicrometer-sized particles through the lters was measured by using a Scanning Mobility Particle Sizer. An Aerodynamic Particle Sizer was used to measure the penetration fraction of aerosol particles larger than 0.8 µ m. The pressure drop across the lter was monitored by using pressure transducers, which were calibrated against an inclined manometer. Air ows of 5, 10 , 20, and 30 cm/s were used to study the ow dependency. The aerosol penetration results showed that the particles larger than 3 µ m did not penetrate the clean fabric lters tested in the present study. The loading curves (plots of pressure drop against sampling time) displayed three regions: an initial region of fast increase, a transition region, and a nal linear region after the dust formation point. After the formation point of the dust cake, both fabric and brous lters shared the same slope (of the loading curves). The slope of different regions of the loading curves was determined by many factors, such as size of challenge aerosol, face velocity, surface treatment, and the compressibility of the dust cake forming on the lter. The method of nal surface treatment was found to be critical to the performance of the fabric lters. In order to avoid the unnecessary rise in air resistance, the melting clumps formed during nal surface treatment should be as thin and narrow as possible, just enough to support the lter bag cleaning. From the standpoint of lter quality and energy consumption, the low ltration velocity has to be adopted whenever possible, because high ltration velocity not only led to lower lter quality (in particular for submicrometer-sized particles) but also created dust cake of lower porosity, which caused an extra rise in pressure drop across the dust cake.