Aims To review post-mortem (PM) examinations performed over a 10 year period and determine if this altered the cause of death or added important additional information.

Methods All babies who died on Birmingham Women's Hospital Neonatal Unit, between January 2000 and December 2009, were identified using a combination of neonatal databases, admissions book and death certificates. Case notes were reviewed and cross-referenced with mortuary records.

Results There were 328 deaths and 91 PMs (82 hospital and 9 coroner's) performed during this period (28%). 10 of the PMs were limited to external examination, biopsy or imaging alone and these were not included in the analysis. 26% of PMs were in term infants, 40% in infants of 28–36 weeks and 33% in babies <28 completed weeks gestation. In 35% of cases there was complete correlation between the PM findings and certified clinical cause of death. In 44% the PM gave additional information that was not immediately related to the cause of death. In 6 babies (8%) the PM gave new information on the cause of death, but knowledge of this prior to death would not have changed the outcome. In 1 case the PM gave important genetic information relevant to future pregnancies.

Conclusion PM remains an important examination both in determining the cause of death and in giving useful additional information. Information gained from a PM is important both for the family, but also in clinical education, research and in monitoring the quality of patient care.