The optimal approach to relatively recent onset type 2 diabetes patients is still unknown. We speculated that the use of short-acting insulin analogs might be of particular benefit in this context. To explore this possibility, we compared the effect on beta- and alpha-cell function of transient intensive insulin therapy using lispro versus human regular insulin in a total of 21 type 2 diabetic patients who were randomly assigned to 14-days intensive insulin therapy consisting of bedtime NPH insulin plus three injections of mealtime lispro (n=11) or regular insulin (n=10). The dosages of both types of insulin were adjusted to attain preprandial glucose levels of <6.1 mmol/l within 1 week with similar rates of glucose decline. An oral glucose tolerance test (OGTT) was performed at day 0 (baseline), 7, and 14; plasma glucose, serum insulin, and plasma glucagon responses over 0-120 minutes were measured, and calculated as the area under the curve (AUC). Lispro led to a significant reduction in glucose-AUC and also an increase in insulin-AUC versus regular insulin on day 7. Glucagon secretion following OGTT was well suppressed with lispro on day 14 compared to regular insulin. Two-week intensive insulin therapy with lispro appeared to be more effective than that with regular insulin in type 2 diabetes in attaining both more rapid beta-cell rest and greater suppression of glucagon. These changes may provide significant long-term benefits.