The performance comparison between different filters requires a test method with repeatable results and low variability. The current SAE J1669 standard does not address some issues related to the measurement of the fractional efficiency of dust loaded filters. This inadequate test procedure can contribute to the unacceptable measurement variability.This paper highlights some of these issues. The influence of media structure, flow rate and amount of dust loading on efficiency measurement is investigated and discussed. The impact of particle shedding on fractional efficiency measurement and variability found to be critical.