Anticholinergic drugs suppress nocturnal and exercise-related growth hormone (GH) secretion in Type 1 diabetes; nocturnal GH suppression is associated with a fall in fasting plasma glucose levels. The aim of this study was to assess the effect of GH suppression on glucose levels following a period of meals and exercise in physiological pattern. Six Type 1 diabetic men recruited from the outpatient clinic were studied in random order at least 1 week apart. After an overnight fast subjects received two-thirds of their usual subcutaneous insulin and either 200 mg oral pirenzepine or placebo at time 0 min. Between 90 and 120 min subjects exercised continuously on an ergometric cycle. Standard meals or snacks were eaten at 30, 150, 270, and 390 min. Venous blood was collected from an indwelling cannula between 0 and 570 min. The mean incremental rise in plasma glucose after breakfast (delta peak/90 min) was 2.6 +/- 0.5 (mean +/- SEM mmol l-1 (pirenzepine) vs 4.5 +/- 0.8 (placebo)), p < 0.05. Following exercise the fall in plasma glucose (delta gluc90-240 min) was 6.4 +/- 1.9 (pirenzepine) vs 2.0 +/- 1.3 (placebo), p < 0.005. The exercise-related peak rise in GH was 12.6 +/- 3.3 (pirenzepine) vs 28.5 +/- 6.0 mU l-1 (placebo), p = 0.08. Excluding one outlying result there was an inverse correlation between the integrated exercise-related increase in GH between 90 and 240 min and the fall in glucose over the corresponding time period (n = 11, r = -0.75, p = 0.008).(ABSTRACT TRUNCATED AT 250 WORDS)