A modle of hypothermic circulatory arrest has been developed in the newborn dog. Ten puppies were anesthetized with halothane, paralyzed, and artificially ventilated with 70% nitrous oxide 30% oxygen to arterial oxygen pressure >8.0 kPa (60 mm Hg), arterial carbon dioxide pressure of 4.4–5.6 kPa (33–42 mm Hg), and arterial pH of 7.35–7.42. Animals were surface cooled to 20°C, after which cardiac arrest was produced with i.v. KCI. Dogs remained asystolic without ventilation for 1.0 (n = 4), 1.5 (n = 3), or 2.0 (n = 3) th Rescucitation was accomplished with closed-chest compression, mechanical ventilation, i.v. epinephrine and NaHCO3, and rewarming to 37°C. Postarrest recovery was maintained for 3–4 h; thereafter, the puppies underwent perfusion-fixation of their brains for pathologic analysis. Plasma glucose (control = 8.3 mmol/L) increased slightly during hypothermic cardiac arrest (+30%) but was markedly elevated at 15 min postarrest (20 mmol/L). Blood lactate (control = 1.1 mmol/L) increased almost 200% during hypothermic circulatory arrest, with a further rise to 9.0 mmol/L at 15 min postarrest. Thereafter, lactate decreased in the 1-h arrested dogs bat increased progressively in the other groups. Mean arterial blood pressure returned to baseline (73 mm Hg) by 15 min postarrest, remained stable in the 1-h dogs, but fell at 3 h to 62 and 34 mm Hg in the 1.5-and 2.0-h groups, respectively. No neuropathologic alterations were seen in puppies arrested for 1 h, whereas all pnppies arrested for 1.5 or 2 h had varying degrees of cerebral cortical and hippocampal damage. Thus, newborn dogs tolerate 1 h of hypothermic circulatory arrest without brain damage, with graded neuronal injury thereafter. The experimental model has direct clinical relevance and can be used to study mechanisms of cellular injory in brain, heart, and other organs during prolonged ischemia. (Pediatr Res28:354–360, 1990)