The aims of this study were to examine the relationship between whole-body absorbed dose and hematologic toxicity and to assess the most accurate method of delivering a prescribed whole-body absorbed dose in (131)I-metaiodobenzylguanidine ((131)I-MIBG) therapy for neuroblastoma. METHODS: A total of 20 children (1-12 y), 5 adolescents (13-17 y), and 1 adult (20 y) with stage 3 or 4 neuroblastoma were treated to a prescribed whole-body absorbed dose, which in most cases was 2 Gy. Forty-eight administrations of (131)I-MIBG were given to the 26 patients, ranging in activity from 1,759 to 32,871 MBq. For 30 administrations, sufficient data were available to assess the effect of whole-body absorbed dose on hematologic toxicity. Comparisons were made between the accuracy with which a whole-body absorbed dose could be predicted using a pretherapy tracer study and the patient's most recent previous therapy. The whole-body absorbed dose that would have been delivered if the administered activity was fixed (7,400 MBq) or determined using a weight-based formula (444 MBq.kg(-1)) was also estimated. RESULTS: The mean whole-body absorbed dose for patients with grade 4 Common Terminology Criteria for Adverse Events (CTCAE) neutropenia after therapy was significantly higher than for those with grade 1 CTCAE neutropenia (1.63 vs. 0.90 Gy; P = 0.05). There was no correlation between administered activity and hematologic toxicity. Absorbed whole-body doses predicted from previous therapies were within +/-10% for 70% of the cases. Fixed-activity administrations gave the largest range in whole-body absorbed dose (0.30-3.11 Gy). CONCLUSION: The results indicate that even in a highly heterogeneous and heavily pretreated patient population, a whole-body absorbed dose can be prescribed accurately and is a more accurate predictor of hematologic toxicity than is administered activity. Therefore, a whole-body absorbed dose can be used to deliver accurate and reproducible (131)I-MIBG therapy on a patient-specific basis.