The need for gasoline particulate filter (GPF) technology is expected to grow with increasingly tight particle emissions standards being implemented in US, EU, China and elsewhere. Derived from the successful experience with diesel particulate filters (DPF), GPFs adopted the characteristic alternately plugged honeycomb structure that provides a large area of porous cordierite wall for filtering particles with minimal additional backpressure. However, unlike DPFs, continuous soot regeneration in GPFs makes it difficult to grow and sustain the soot cake on the filter wall that gives DPFs their high filtration efficiency. Therefore, filtration performance of low mileage GPFs relies heavily on the porous structure of filter media, which depends on both the substrate and the applied washcoat. In this work, a blank, two fresh washcoated filters and two washcoated filters with 3000 km mileage accumulation were characterized to compare their filtration performance. Particle number and mass filtration efficiency and pressure losses of filters with and without soot loading were measured on an aerosol test bench, and the results are compared with findings from engine dynamometer tests. Compared with blank filters, washcoated filters in this study generally had lower filtration efficiency and higher backpressure when clean, but efficiency increased when loaded with small amounts of soot (~20 mg/L or less). Ash collected during mileage accumulation improved the efficiency significantly, with only a mild backpressure increase. The mechanisms leading to the observed trends in filtration efficiency and backpressure are discussed.