OBJECTIVE: The HypoCOMPaSS study was designed to test the hypothesis that successful avoidance of biochemical hypoglycemia without compromising overall glycemic control would restore sufficient hypoglycemia awareness to prevent recurrent severe hypoglycemia in the majority of participants with established type 1 diabetes. Before starting the study, we planned to investigate associations between baseline characteristics and recurrent severe hypoglycemia over 2 years' follow-up. RESEARCH DESIGN AND METHODS: A total of 96 adults with type 1 diabetes and impaired awareness of hypoglycemia participated in a 24-week 2 × 2 factorial randomized controlled trial comparing insulin delivery and glucose monitoring modalities, with the goal of rigorous biochemical hypoglycemia avoidance. The analysis included 71 participants who had experienced severe hypoglycemia in the 12-month prestudy with confirmed absence (complete responder) or presence (incomplete responder) of severe hypoglycemia over 24 months' follow-up. RESULTS: There were 43 (61%) complete responders and 28 (39%) incomplete responders experiencing mean ± SD 1.5 ± 1.0 severe hypoglycemia events/person-year. At 24 months, incomplete responders spent no more time with glucose ≤3 mmol/L (1.4 ± 2.1% vs. 3.0 ± 4.8% for complete responders; P = 0.26), with lower total daily insulin dose (0.45 vs. 0.58 units/24 h; P = 0.01) and greater impairment of hypoglycemia awareness (Clarke score: 3.8 ± 2.2 vs. 2.0 ± 1.9; P = 0.01). Baseline severe hypoglycemia rate (16.9 ± 16.3 vs. 6.4 ± 10.8 events/person-year; P = 0.002) and fear of hypoglycemia were higher in incomplete responders. Peripheral neuropathy was more prevalent in incomplete responders (11 [39%] vs. 2 [4.7%]; P < 0.001) with a trend toward increased autonomic neuropathy. CONCLUSIONS: Recurrent severe hypoglycemia was associated with higher preintervention severe hypoglycemia rate, fear of hypoglycemia, and concomitant neuropathy.