Blood glucose (BG) response to psychological stress in insulin-dependent diabetes mellitus (IDDM) patients has not been firmly established. We report a study designed to address the gaps and methodological difficulties reviewed. Subjects with IDDM were exposed to two sessions (12 weeks apart) of two 20-min standardized stressors (active and passive) and a control condition administered in counterbalanced order. To measure BG response, subjects were connected to a glucose/insulin infusion system providing continuous BG measurement. Mood checklist measures were obtained at prestressor, poststressor, and recovery periods. During the first session of testing, the active stressor was associated with significantly more absolute change in BG response than the passive stressor. Results also indicate that IDDM subjects' BG response to this active stressor was idiosyncratic but significantly reliable over time.