beta-Thromboglobulin (beta-TG), which is a specific platelet protein, is considered to be a fair indication of in vivo platelet activation. Our investigation was aimed at testing the influence of the glucose balance on platelet behaviour by measuring beta-TG levels in insulin-dependent diabetic patients subjected to blood feed-back control via an artificial pancreas over a 24 or 48 hour period. In 19 patients who underwent feed-back control over a 24 hour period, no significant change in beta-TG level was observed on average, compared with the measurements carried out before blood glucose control. In contrast, in the 11 diabetic patients treated over a 48 hour period, this plasma level decreases slightly but significantly compared with the levels observed before glycemia regulation (55.6 +/- 26.6 ng/ml vs. 74.2 +/- 24.9 ng/ml). We were unable to find any correlation in these patients between the beta-TG levels measured at each of the three sampling periods and the blood glucose level-measured simultaneously or continuously-or the glycosylated hemoglobin level measured at the beginning of the study. These results suggest that, in spite of the lag phase, blood glucose control has a beneficial influence on the platelet hyperactivity and therefore a poor glycemia regulation could be largely responsible of this disorder. However, in view of the persistence of abnormally high plasma beta-TG level after 48 hours, we are led to believe that strict blood glucose control has not been sufficiently prolonged.