I n 1941, writing in the Lancet, R. D. Lawrence (1) noted that " [hypoglycemic] reactions differ so much from the original ones that patients become dangerously unaware of their onset." This clinical description of impaired hypoglycemia awareness is still recognizable today, and we have become increasingly alert to its development after the publication of the landmark Diabetes Control and Complications Trial (DCCT) in 1993. The DCCT demonstrated that good glycemic control with intensive insulin therapy markedly reduced the risk of long-term microvascular complications of type 1 diabetes (T1D) (2) but that this came at the expense of a significant increase in the risk of biochemical and symptomatic hypoglycemia, whereas the risk of severe hypoglycemia increased 3-fold (3). Intensive insulin therapy was shown to reduce the glucose level at which hypoglycemic counterregulatory and symptomatic defense responses were initiated, significantly contributing to an individual's risk of severe hypoglycemia (4). A number of studies in humans and animals then showed convincingly that prior exposure to hypoglycemia increased the threshold (lower glucose level) for hormone and symptom responses to subsequent hypoglycemia, which when combined with other counterregulatory defects in T1D (primarily the loss of pancreatic ␣-cell glucagon secretion) was called the syndrome of hypoglycemia-associated autonomic failure (HAAF) (5). Studies, largely in animal models, have since helped inform us about how and where hypoglycemia is sensed in the body (6). However, as pointed out by Amiel (7) in a recent review, our ability to actually prevent the development of HAAF other than through strenuous efforts to avoid hypoglycemia is very limited, more often than not just leading to a period of more relaxed glucose control.In light of this, the report by Vele et al. (8) in this issue of the JCEM, in which the nonspecific opioid receptor antagonist naloxone prevented the development of HAAF in individuals with intensively treated T1D who were studied using an established experimental model, is an important one. Eight subjects with T1D (age, 34 Ϯ 7 yr; duration of diabetes, 12 Ϯ 3 yr), all treated with insulin pumps and with good glycemic control (glycosylated hemoglobin, 7.3 Ϯ 1.1%), took part in three different studies in random order and each separated by at least 5 wk. Each study required the subject to attend on 2 consecutive days. On d 1, each subject underwent two 90-min hyperinsulinemic glucose clamp studies separated by 60 min. The conditions induced were euglycemia (90 mg/dl) or hypoglycemia (60 mg/dl), the latter with (Nϩ) or without (NϪ) a concomitant infusion of naloxone. On d 2, all subjects underwent a stepped (90, 80, 70, and 60 mg/dl) hypoglycemic clamp study where each glucose plateau was maintained for 50 min, and blood samples were taken for measurement of counterregulatory hormones and glucose turnover. The investigators found as expected that two episodes of antecedent hypoglycemia (control euglycemia vs. NϪ) resulted in a ...