Seventy children (3.7 +/- 3.3 y) with definitely confirmed diagnosis of neuroblastoma had 115 whole body scans carried out 24 h after injection of 3.7 MBq/kg of I-123 mIBG (83 scans) or 0.7 MBq/kg of I-131 mIBG (17 scans) or 0.9 to 4.5 GBq of I-131 mIBG (15 post-therapeutic scans). The scans were interpreted as positive in the presence of any non-physiological uptake area or of any bone uptake of the tracer, even at the level of the metaphyseal complex. For the primary tumour, the sensitivity of mIBG scans was 73%. Ten false negative patients had an overlap of the tumour with the bladder or heart images (4 cases) or with positive metastatic images (6 cases: liver, spine). Three false negative patients had neuroblastomas which did not secrete catecholamines. The specificity of mIBG was 94%. In our opinion, mIBG scans have a complementary role to assess the activity of post-therapeutic remnants. For the detection of hepatic and lymph node metastases, the sensitivity was about 50% and the specificity was 100%. The standard used for the detection of bone marrow metastases was the cytological and histological examination of 10 bone marrow aspirations and one or more biopsies (CHBMS). The sensitivity of mIBG scans was 90% and the specificity 68%. However, reviewing the data from the 16 false positive scans, we found 11 definitely proven bone metastases, 3 biological relapses and 2 cases of delayed abnormal CHBMS supporting the positivity of the mIBG scans, raising the specificity to 100%. Tc-99m diphosphonate bone scans had a sensitivity of 78% and a specificity of 51%. We suggest that positive mIBG scans may save other procedures since our data do not support false positive detection of bone or bone marrow metastases. In contrast, patients with negative mIBG findings should be further explored.