Aims: This study set out to define relationships between changes in plasma leptin and changes in body weight, plasma insulin and blood glucose control during a 12-month crossover study of once-daily Ultratard or twice-daily Insulatard insulin. Patients and Methods: Fasting plasma leptin and insulin were measured during a multicentre cross-over study involving 60 subjects with type 2 diabetes (fasting glucose > 8 mM). After a 2-month run-in, there were two 6-month periods of treatment with Insulatard or Ultratard insulin. Results: Mean plasma leptin increased significantly in both groups after insulin therapy was instigated (12.8 ± 8.1 to 22.9 ± 13.1 ng/ml in the Insulatard group; 12.1 ± 7.2 to 19.2 ± 12.3 ng/ml in the Ultratard group). Weight also increased significantly in both groups (82.4 ± 14.3 kg to 88.8 ± 14.3 kg and 82.2 ± 15.3 kg to 85.3 ± 15.2 kg respectively). The increase in plasma leptin correlated well with the increase in weight (R = 0.416, p = 0.001), and this correlation continued after the crossover point. Plasma leptin correlated with BMI throughout the study (R = 0.540, p = 0.000). Conclusion: The sustained rise in body weight despite a substantial increase in plasma leptin suggests that either resistance to the hypothalamic action of leptin develops when insulin therapy is begun in type 2 diabetes, or that resetting of the set point for body weight occurs such that a larger body mass is tolerated for a given level of plasma leptin.