The measurement of low frequency impact sound in rooms (below 100 Hz) is critical to determining human response and acceptability of floor ceilings [1], [2]. Measurement uncertainties generally increase at lower frequencies, and a lack of precision is a general problem that ratings of low-frequency noise must overcome. Impact noise uncertainties at low frequencies are reviewed. The effect on measurement uncertainty of changes to the measurement procedure (such as fixed vs. roving microphones, measurement duration, microphone position, number of tapping machine locations, etc.) is studied, with emphasis on the uncertainties in the 50-80 Hz third-octave bands. The measurement uncertainty will be related to the precision desired to accurately relate to human reaction.