Type 2 (non-insulin-dependent) diabetes mellitus is the major form of the disease in all societies. Its public health impact appears to be increasing and the greatest genetic predisposition to the disease is encountered in developing communities. The reduction or elimination of disease in whole populations is a fundamental goal in public health. Whilst several factors are associated with the development of Type 2 diabetes, it is not clear how they cause the disease, if indeed they do, nor whether they act in the same way in all populations. Risk factors may be true determinants of a disease but alternatively they may be associated with its occurrence only by virtue of an innocent relationship with the true causes. Furthermore, known risk factors usually explain only a small proportion of any chronic disease. The role of risk factors in disease causation is therefore of fundamental importance in considering disease prevention. Two alternative strategies for prevention of disease in populations have been proposed. The population strategy seeks to remove the causes of disease in communities as a whole, whilst the high-risk strategy aims to identify subjects at increased risk, and to intervene selectively. The population approach should be tried and carefully evaluated in selected communities at above-average risk of several noncommunicable diseases. However, certain epidemiological features of Type 2 diabetes, including the distributional characteristics of glycaemia and the complications of hyperglycaemia, the clustering of cardiovascular risk factors in the diabetic subpopulation, as well as uncertainties over the causal nature of known risk factors, suggest that a high-risk approach to prevention is also appropriate.(ABSTRACT TRUNCATED AT 250 WORDS)