Diet in Type 2 diabetes often reduces both hyperglycaemia and weight and both these factors may contribute to improvement in beta cell function. Sixty-nine subjects with newly diagnosed Type 2 diabetes had their glucose, insulin, intact and 32/33 split proinsulin measured at diagnosis, and after 16 (12-20) weeks of conventional diet. In the whole group following diet there was reduction in weight (p < 0.0001), Haemoglobin A1 (p < 0.0001), and fasting glucose (p < 0.0001). There was a fall in fasting intact proinsulin (p < 0.03), 32/33 split proinsulin (p < 0.0001), and the percentage of fasting proinsulin to total insulin-like molecules (p < 0.0001). Subjects were separated according to weight loss (group 1 < 0.5 kg, group 2 0.5-4 kg, group 3 > 4 kg). Only subjects in group 3 (mean weight loss 6.2 kg) had a fall in fasting insulin (p < 0.04). The fasting glucose achieved following diet was positively correlated with the initial fasting glucose (r = 0.7) and negatively correlated with the degree of insulin deficiency at diagnosis (with the initial 30 min insulin r = -0.62). The effect of diet on the final concentration of insulin, 32/33 split proinsulin and intact proinsulin was examined using multiple regression. The final insulin concentration was determined by the initial BMI and absolute weight loss, 32/33 split proinsulin concentration by the initial fasting glucose and absolute glucose fall, and intact proinsulin concentration by a combination of both weight loss and glucose reduction.(ABSTRACT TRUNCATED AT 250 WORDS)