Aims/hypothesis. We have previously shown that lactate protects brain function during insulin-induced hypoglycaemia. An adaptation process could, however, not be excluded because the blood lactate increase preceded hypoglycaemia.¶Methods. We studied seven healthy volunteers and seven patients with Type I (insulin-dependent) diabetes mellitus with a hyperinsulinaemic (1.5 mU · kg^–1· min^–1) stepwise hypoglycaemic clamp (4.8 to 3.6, 3.0 and 2.8 mmo/l) with and without Na-lactate infusion (30 μmol · kg^–1· min^–1) given after initiation of hypoglycaemic symptoms.¶Results. The glucose threshold for epinephrine response was similar (control subjects 3.2 ± 0.1 vs 3.2 ± 0.1, diabetic patients = 3.5 ± 0.1 vs 3.5 ± 0.1 mmol/l) in both studies. The magnitude of the response was, however, blunted by lactate infusion (AUC; control subjects 65 ± 28 vs 314 ± 55 nmol/l/180 min, zenith = 2.6 ± 0.5 vs 4.8 ± 0.7 nmol/l, p < 0.05; diabetic patients = 102 ± 14 vs 205 ± 40 nmol/l/180 min, zenith = 1.4 ± 0.4 vs 3.2 ± 0.3 nmol/l, p < 0.01). The glucose threshold for symptoms was also similar (C = autonomic 3.0 ± 0.1 vs 3.0 ± 0.1, neuroglycopenic = 2.8 ± 0.1 vs 2.9 ± 0.1 mmol/l, D = autonomic 3.2 ± 0.1 vs 3.2 ± 0.1, neuroglycopenic 3.1 ± 0.1 vs 3.2 ± 0.1 mmol/l) but peak responses were significantly attenuated by lactate (score at 160 min C = 2.6 ± 1 vs 8.8 ± 1, and 0.4 ± 0.4 vs 4.8 ± 1, respectively; p = 0.02–0.01, D = 1.3 ± 0.5 vs 6.3 ± 1.7, and 2.3 ± 0.6 vs 5.7 ± 1.1 p = 0.07–0.02). Cognitive function deteriorated in both studies at similar glucose thresholds (C = 3.1 ± 0.1 vs 3.0 ± 0.1, D = 3.2 ± 0.1 vs 3.3 ± 0.2 mmol/l). Although in normal subjects a much smaller impairment was observed with lactate infusion (Δ four-choice reaction time at 160 min = 22 ± 12 vs 77 ± 31 ms; p = 0.02), in Type I diabetic patients lactate infusion was associated with an improvement in cognitive dysfunction (0.2 ± 0.4 vs –38 ± 0.2 Δ ms, p = 0.0001).¶Conclusion/interpretation. A blood lactate increase after the development of hypoglycaemic symptoms reduces counterregulatory and symptomatic responses to insulin-induced hypoglycaemia and favours brain function rescue both in normal and diabetic subjects. These findings confirm that lactate is an alternative substrate to glucose for cerebral metabolism under hypoglycaemic conditions. [Diabetologia (2000) 43: 733–741]