In earlier parts of this series of papers on the prediction of impact noise, it has been found that in predicting the noise energy radiated from an industrial machine, the only term in the energy accountancy equation which involves the true conversion of vibrational energy into heat is the quantity 10 log ηs; the other terms represent the fraction of impact energy entering the machine and the radiation efficiency change associated with moving this vibrational energy to lower frequencies. Thus the study of the overall damping factor ηs is of crucial importance to the accurate prediction of noise radiated. In spite of the large bibliography available on damping, the practical prediction of this quantity in industrial type machinery is so uncertain that many workers treat the quantity as an unknown "fudge factor" to be obtained from previous similar machines. This forbids the deliberate "designing in" of damping in a new machine, and leads to disappointment if new practices have inadvertently caused a significant loss in ηs, especially when, in fact, the previous versions were relatively highly damped. In this paper a study aimed at improving damping prediction is described. Based upon an investigation of the values of ηs obtained in industrial machinery structures, as opposed to "thin shell" viscoelastically damped structures, a review is presented of the levels of damping which can be obtained by various standard methods. The effects of bolts and fluid sloshing are included, and specific experiments are described on the effects of adding aggregates in cavities, adding close covers and fitting stick-slip springs on drill rods. There is ample evidence that adequate damping may be obtainable only by the additi·n of several of these palliatives to different parts of the machinery structure, and accordingly a possible method of summation is proposed, based upon an analogy with room acoustics. The study has led to a realization of the importance of obtaining a simple method of summating damping, and further work is now being done to validate such a method.