OBJECTIVE: To examine the influence of caffeine on the frequency and perception of hypoglycemia in "free-living" patients with type 1 diabetes. RESEARCH DESIGN AND METHODS: A total of 34 patients with type 1 diabetes were recruited for a prospective randomized placebo-controlled double-blind study. After a lead-in phase and while adhering to a low-caffeine diet, subjects were randomized to capsules containing either 200 mg caffeine or matched placebo with crossover at 3 months. Hypoglycemic episodes were monitored throughout with capillary blood glucose readings and a symptom questionnaire. During the study, measurements of blood pressure, middle cerebral artery blood velocity (a surrogate measure of cerebral blood flow), cognitive function (via a four-choice reaction time test), HbAlc levels, and lipid profiles were taken at the beginning and end of each phase. RESULTS: Throughout the study, no changes were evident regarding glycemic control or lipid profile. The number of symptomatic episodes was greater with caffeine (1.3 vs. 0.9 episodes/week; P < 0.03) and was associated with more intense warning symptoms (29 vs. 26 total symptom score; P < 0.05). For women, caffeine ingestion caused a modest pressor response (115 vs. 110 mmHg; P < 0.01). Four-choice reaction time improved slightly with caffeine supplementation (P < 0.05). CONCLUSIONS: Ingestion of modest amounts of caffeine enhances the intensity of hypoglycemia warning symptoms in patients with type 1 diabetes without altering the prevailing standard of glycemic control or increasing the incidence of severe hypoglycemic episodes.