This review concentrates on the effects of low frequency noise (LFN) up to 100 Hz on selected physiological parameters, subjective complaints and performance. The results of laboratory experiments and field studies are discussed in relation to the thresholds of hearing, of vibrotactile sensation and of aural pain. The effects of LFN may be mediated trough different ways. Temporary or permanent hearing threshold shifts seem to be due to acoustic stimuli above the individual hearing threshold. However, non-aural physiological and psychological effects may be caused by levels of low frequency noise below the individual hearing threshold. The dynamic range between the thresholds of hearing and of aural pain diminishes with decreasing frequency. This should be taken into account by the setting of limits concerning the health risks. Sufficient safety margins are recommended. The use of a frequency weighting with an attenuation of the low frequencies (e.g. G-weighting) does not seem to be appropriate for the evaluation of the health risks caused by LFN up to 100 Hz. It may be proposed to measure third octave band spectra or narrow band spectra. A comparison with the known human responses caused by the measured levels and frequencies could help to evaluate the health risks. Some proposals for further investigations were given: (1) experimental methods to discover the ways mediating the effects of low frequency noise, (2) consideration of the individual hearing threshold or hearing threshold shift and of the vibrotactile threshold in the low frequency range to be able to judge the effects, (3) consideration of combined body vibration caused by airborne low frequency noise or by other sources, (4) modelling to analyse the transmission of the acoustic energy from the input into the body to the structures containing sensors, (5) consideration of probable risk groups like children or pregnant women.