Objective:Information regarding cerebral oedema in diabetic children with ketoacidosis does not point to any causal factor or any predictor of outcome. Cases of diabetic ketoacidosis resulting in cerebral oedema at Royal Children's Hospital, Melbourne, over the last 20 years were reviewed.Methodology:The study was divided into two 10‐year periods during which different fluid protocols were used. During the earlier period dehydration was corrected rapidly (over six h) with a fluid isotonic for sodium. During the latter period rehydration was over 24 h using half‐normal saline.Results:A similar number of patients developed cerebral oedema in the two periods (six of 3134 vs six of 3373). Only half of the patients (six) developed cerebral oedema during their initial presentation of diabetes mellitus. The age range was similar (1 to 15 years) with eight males and four females. Survival from cerebral oedema (four of 12) was not predicted by demographic or biochemical findings, including initial biochemistry, age, duration of ketoacidosis and the management protocols.Conclusions:This study suggests that the rate of salt and water replacement in diabetic ketoacidosis are not key determinants of the appearance of cerebral oedema. No factors predictive of survival from cerebral oedema have been identified, though this is a rare entity and case numbers were small. Nevertheless, current protocols at Royal Children's Hospital and most other centres utilize slow rates of rehydration with isotonic saline fluids. Further review in 5–10 years may determine whether this protocol is effective in reducing rates of cerebral oedema complicating diabetic ketoaci