Our study is based on two constatations: 1) Hyperinsulinaemia, a possible atherogenic factor, is frequent under continuous subcutaneous insulin infusion. 2) Pulsatile intravenous insulin delivery improve the insulin's hypoglycaemic activity. To test if equivalent metabolic control can be obtained with a reduced intermittent subcutaneous infused insulin dose, we compared nocturnal metabolic control of 8 c-peptide negative type 1 diabetic patients under three experimental conditions: Continuous usual dose test (1.0 +/- 0.1 u/h); Intermittent half dose test (1.0 +/- 0.1 u/h, 30 min/h); Continuous half dose test (0.5 +/- 0.05 u/h) Five parameters were monitored: blood glucose, plasma free insulin and beta-hydroxy-butyrate, free fatty acid and glycerol plasma level. No significant differences were found between intermittent and continuous half-dose tests. We conclude that, in our experimental conditions, intermittent subcutaneous insulin infusion does not reduce the metabolic degradation induced by insulin dose reduction.