The glucose clamp technique has been used to evaluate the metabolic activity of NPH biosynthetic insulin in diabetic subjects free from anti-insulin antibodies. After overnight blood glucose normalization with a glucose-controlled insulin infusion system (Biostator), an s.c. injection of NPH insulin was given in the abdominal region. The insulin dose (0.236 +/- 0.05 U/kg body wt) was related to the usual intermediate-acting insulin requirement in the morning. Glucose was clamped at 100 mg/dl by a feedback i.v. glucose infusion. The end of the action of s.c. injected insulin considered conventionally to be the time of the spontaneous rise of blood glucose to 110 mg/dl. Free insulin levels were higher and the length of action was longer after NPH porcine than after NPH biosynthetic human insulin (BHI) (area under the free insulin curve: porcine 1423 +/- 556 mU/L/h; BHI 1045 +/- 338 mU/L/h, P less than 0.05; length of action: porcine 16.0 +/- 3.2 h; BHI 13.7 +/- 0.9 h, P less than 0.05); the glucose requirement was higher after porcine (76.8 +/- 13.5 g) than after BHI (58.5 +/- 14.6 g) without reaching statistical significance. However, the metabolic activity of the bioavailable insulin (index of plasma free insulin activity) was similar for the two insulins (porcine 381 +/- 77.4, BHI 342.8 +/- 54.2 mU/L/g of glucose/h). We conclude that a difference in pharmacokinetics exists between NPH BHI and porcine NPH insulin, which makes the latter metabolically more active. The different behavior does not seem to be related to the insulin molecule itself but could be a consequence of the unequal content of protamine in the two pharmacologic preparations.