Severe noise problems involving a small number of people can often be solved following a detailed examination of the factors involved. At very least, provided the necessary resources are available, the noise source and the offended people can usually be separated by distance or by sound insulation. More usually, however, real noise nuisance problems are very complex, involving trade-offs between the benefits of less noise and the costs of lessening it. It is clearly desirable to minimise the harmful effects of noise upon public health but how low must the limits be set and can the costs be met? In general the environmental planner is faced with two problems, firstly to predict the extent and severity of environmental noise impact and secondly to assess the acceptability, or otherwise, of this impact in some absolute sense. The second of these problems is perhaps the more difficult as its solution ultimately rests on the political question of how many affected people can be justified. But the first is of no less importance since the policy-maker must be able to make qualitative evaluations of the effects of different policy options. This paper examines the problem of environmental noise nuisance what it is and how it can be measured and predicted. Detailed methodology is beyond the scope of the paper which concentrates instead upon a discussion of the factors involved and the nature of the relationships between them.