Major genes for NIDDM appear to be rare. Therefore, phenotypic characterization of the pathophysiological changes contributing to hyperglycaemia are assuming increasing importance. Assessment of beta-cell function has been hampered by two major confounding factors during functional testing: variable insulin sensitivity and plasma glucose levels. These and other methodological variables are discussed with recommendations for ameliorating or accounting for their impact. A group of tests as used by the Seattle Group for phenotypic characterization is described including basal immunoreactive insulin (IRI), proinsulin, and proinsulin intermediates (PI); the acute insulin response to glucose (AIRg); maximal (AIRmax), and half-maximal (PG50) capacity to potentiate a non-glucose secretagogue; and insulin-sensitivity (S1). Specific examples in the use of this battery of tests are given. It is concluded that such phenotyping will be an important tool for studies of the epidemiology and genetics of NIDDM.