Nineteen patients suffering from chemical diabetes either with (group A, ten cases) or without (group B, nine cases) reactive hypoglycaemia were included in the study and compared with seven control (group C). The following variables were measured over a 5 hour period during a standard oral glucose tolerance test (OGTT): (i) blood glucose by continuous monitoring; (ii) plasma insulin and glucagon levels by radioimmunoassay. Furthermore, in five diabetics of group A, the data from the standard OGTT were compared with those from a pectin-supplemented OGTT (9 g per square meter of body surface). Although the insulin response was similar glucagon levels were significantly higher (45.1 +/- 11.8 pmol/l) (p less than 0.01) in group B than in group A (9.6 +/- 1.3) and C (8.1 +/- 1.4 at 30 minutes). The high glucagon levels noted in group B may explain the absence of reactive hypoglycaemia. The pectin supplementation improved the OGTT pattern by blunting the blood glucose peak (p less than 0.05), and avoiding the reactive hypoglycaemia (p less than 0.01). The addition of pectin did not produce any significant effect on the insulin response while a significant increase in glucagon concentrations (p less than 0.05) was observed beyond the 150th minute. Therefore, the data suggest that pectin may improve the OGTT pattern by increasing the glucagon response in the late period of the test. The development of postprandial reactive hypoglycaemia seldom coincides with a plasma glucagon peak, while the absence of reactive hypoglycaemia tends to be associated with high levels of glucagon, as is the case in overt diabetes mellitus.