Continuous intraperitoneal insulin infusion (CIPII) is a promising therapy of patients with Type 1 (insulin-dependent) diabetes mellitus (IDDM), since it improves metabolic control and decreases frequency of severe hypoglycaemia. This could be due to more appropriate insulin kinetics. Our aim, therefore, was to compare plasma free insulin levels achieved in patients with Type 1 diabetes chronically treated with CSII or CIPII. Furthermore, as anti-insulin antibodies increase with this treatment, we wanted to assess their influence upon insulin kinetics. Plasma free insulin profiles were obtained during the night and then after the bolus for breakfast and the bolus for lunch in 11 patients with Type 1 diabetes treated successively by CSII and CIPII. In another group of 16 patients with long-term Type 1 diabetes, treated by CIPII, we examined the influence of anti-insulin antibody level on insulin kinetics after a bolus. During the night, plasma free insulin levels were lower with CIPII than with CSII (12:00 am: 10.1 +/- 1.7 vs 18.5 +/- 2.6 mU l-1; 4:00 am: 9.1 +/- 2 vs 15 +/- 3 mU l-1), p < 0.01. After the bolus, CIPII lead to an earlier (1h vs 3h) and higher (25.8 +/- 3.3 vs 18 +/- 2.7, p < 0.05) plasma free insulin peak than CSII. With CIPII, the return to baseline level was observed within 3 h. Conversely, during CSII, insulin levels did not return to baseline until the next meal. After the bolus, high insulin-antibody levels were associated with a reduced maximal value of plasma free insulin peak. Taken together, these findings suggest that CIPII provides plasma free insulin profiles which are much closer to physiology than CSII. This could explain the lower rate of severe hypoglycaemia observed with this type of treatment. But in long-term CIPII treated patients with high anti-insulin antibody level, insulin profile could be moderately modified. This emphasizes the need for a less immunogenic insulin preparation.