The World Health Organization distinguishes among non-diabetic persons a form of glucose intolerance defined as impaired glucose tolerance (IGT). The main reason for considering IGT as a diagnostic entity is its prognostic value for the development of non-insulin-dependent diabetes mellitus (NIDDM). However, the use of one or two oral glucose tolerance tests (OGTT) for the definition of IGT and the large variability of the 2-h glucose level may explain the wide range of the incidence rates reported for NIDDM in subjects with IGT. It is evident that the pathogenesis of diabetes is still poorly understood. Both insulin resistance and impaired function of the beta-cell are thought to be important contributing factors in the development of diabetes. The 2-h post-load glucose level from the OGTT is most powerful for the prediction of the development of NIDDM. In addition, a fasting hyperinsulinaemia and a low glucose removal rate, both reflecting insulin resistance, were found to be associated with a higher risk for the progression of NIDDM. In the San Antonio Study and in the Hoorn Study high fasting proinsulin levels, reflecting beta-cell dysfunction, were associated with progression to NIDDM. Lipid levels did not predict progression to NIDDM in most studies. A two-step model for the development of NIDDM is hypothesized. The first step, the transition from normal to impaired glucose tolerance, is assumed to depend mainly on the presence of insulin resistance. The second step, progressing from IGT to diabetes, although accompanied by some further worsening of insulin resistance, is thought to be primarily dependent on the development of beta-cell dysfunction.