On the basis of the available data (much of which is contradictory), I suggest that the following might summarize the role of Lp(a) in diabetes currently. 1. Lp(a) in IDDM: Concentrations are probably elevated. Concentrations are probably related to metabolic control. Concentrations are increased with microalbuminuria. 2. Lp(a) in NIDDM: Concentrations are not elevated. Concentrations do not change with metabolic control. Too few data exist to make an assessment of relation of Lp(a) to microalbuminuria in NIDDM. 3. Lp(a) and CHD in diabetes: Little current evidence shows that Lp(a) is a risk factor for CHD in diabetes. More studies--especially prospective studies with larger numbers of subjects--need to be done.