A double-blind, placebo-controlled investigation has been made into the effects of 8 weeks of glipizide treatment in diabetics previously classified as Type 2 but with subsequent attenuation of insulin secretion and thence maintained on exogenous insulin. Although all patients were exposed to therapeutic plasma concentrations of glipizide, fasting blood glucose, haemoglobin A1 and plasma lipoproteins (HDL, LDL, total cholesterol and triglycerides) did not show any consistent improvement following this treatment. It appears unlikely that SU (glipizide) has any primary effect on insulin action or on plasma lipoproteins. Its primary action is to augment insulin release and availability, so, its use should be restricted to Type 2 diabetics who retain insulin secretion.