We devised a diagnosis and management algorithm for acute onset of central diabetes insipidus (CDI), and conducted a retrospective evaluation of its efficacy. Fourteen patients admitted to our pediatric intensive care unit (PICU) over a three year period were diagnosed with acute CDI secondary to various brain injuries. All patients were treated as per the algorithm guidelines. The initial dose of aqueous vasopressin ranged from 0.25 to 1.0 mU/kg/h. Low sodium content solution (0-0.5 normal saline) was used to replace urine output in excess of 3 ml/kg/h and for maintenance fluid therapy. The therapeutic goals included: urine output 2-3 ml/kg/h, urine specific gravity 1.010-1.020 and serum sodium 140-145 mEq/l. The pitressin dose was adjusted as deemed necessary to achieve the aforementioned goals. Our results indicate that urine specific gravity is the most sensitive parameter to respond to treatment. It was the best determinant of the adequacy of pitressin dose as it had the best linear correlation with it (r = 0.96; p = 0.009). Urine output was second best (r = 0.93; p = 0.02), whereas no linear correlation was established between pitressin dose and serum sodium concentration, nor with serum osmolality. We conclude that the algorithm developed and used by us for the management of CDI is generally efficacious. Changes in urine specific gravity follow changes in pitressin dose very closely and thus should be used as the primary parameter for determination of intravenous pitressin dose adjustment.