The aim of this thesis was to examine the validity of three different theoretical models in explaining the nature of alcohol-related problems in both clinical and general populations. The two models which have traditionally occupied centre stage in the public health debate concerning the most appropriate means to reduce the burden of alcohol-related problems on society, are described here as the 'cluster' and 'disaggregation' models. The former has often, although not exclusively, been described in terms of a more medically-based disease concept, whereas the latter represents more a sociologically-based concept favoured by the epidemiologist. The cluster model advises a public health policy of specialized treatment targetted at the minority of very heavy drinkers, whereas the disaggregation model suggests alcohol control policies such as taxation aimed at reducing alcohol consumption in the whole drinking population.