835 ,umol 1-1 h, AUC, P = NS) were not significant, but the peak increments of acetoacetate (1413 ± 354 ,umol F-1 vs 612 ± 176 ,mol F-1, P < 0.05), ,3-hydroxybutyrate (2180 ± 475 ,umol 1-1 vs 922 ± 246 ,umol 1-1, P < 0.01) and the decrements in plasma bicarbonate (-8 ± 1 ,umol F-1 vs -4 ± 1 ,umol F-1, P < 0.05) and pH (-0.07 + 0.01 vs -0.03 ± 0.01, P < 0.05) were significantly less with octreotide. 4 At the same time peak increments of glucagon were lower with octreotide treatment (329 ± 206 pg ml-' vs 39 ± 30 pg ml-', P < 0.05). 5 We conclude that, despite accelerated lipolysis and provision of substrate for ketogenesis during insulin withdrawal, this somatostatin analogue significantly reduces ketogenesis resulting from insulin deprivation, probably secondary to decreasing glucagon secretion. This drug may be useful in short term prophylactic treatment of diabetic patients during periods of increased risk for ketoacidosis.