BACKGROUND: Cerebrovascular reactivity monitoring has been used to identify the lower limit of pressure autoregulation in adult patients with brain injury. We hypothesise that impaired cerebrovascular reactivity and time spent below the lower limit of autoregulation during cardiopulmonary bypass will result in hypoperfusion injuries to the brain detectable by elevation in serum glial fibrillary acidic protein level. METHODS: We designed a multicentre observational pilot study combining concurrent cerebrovascular reactivity and biomarker monitoring during cardiopulmonary bypass. All children undergoing bypass for CHD were eligible. Autoregulation was monitored with the haemoglobin volume index, a moving correlation coefficient between the mean arterial blood pressure and the near-infrared spectroscopy-based trend of cerebral blood volume. Both haemoglobin volume index and glial fibrillary acidic protein data were analysed by phases of bypass. Each patient's autoregulation curve was analysed to identify the lower limit of autoregulation and optimal arterial blood pressure. RESULTS: A total of 57 children had autoregulation and biomarker data for all phases of bypass. The mean baseline haemoglobin volume index was 0.084. Haemoglobin volume index increased with lowering of pressure with 82% demonstrating a lower limit of autoregulation (41+/-9 mmHg), whereas 100% demonstrated optimal blood pressure (48+/-11 mmHg). There was a significant association between an individual's peak autoregulation and biomarker values (p=0.01). CONCLUSIONS: Individual, dynamic non-invasive cerebrovascular reactivity monitoring demonstrated transient periods of impairment related to possible silent brain injury. The association between an impaired autoregulation burden and elevation in the serum brain biomarker may identify brain perfusion risk that could result in injury.