In order to achieve some uniformity in histological detection of bone marrow infiltration by neuroblastoma and to provide a measure of variation in histological opinions, sections from 712 evaluable trephine biopsy cores from children in a European Neuroblastoma Study Group (ENSG) study were reviewed centrally. Biopsy specimens were graded as tumour positive or negative. Discordance between local and central review opinions was found in 5% of specimens. Only five of 165 children at presentation and nine of 256 re-staging procedures in 126 children, affecting one child each, had their diagnosis upgraded to positive. In six re-staging procedures, affecting one child each, the diagnosis was downgraded. The low discordance rate is encouraging and substantially less important than previously documented difficulties in obtaining adequate specimens.