Spontaneous tumour regression is a well-recognised phenomenon in infantile favourable-biology neuroblastoma. An ‘expectant-observation’ strategy avoids chemotherapy or surgery associated risks but has mostly been limited to small tumours (diameter < 5 cm) and discontinued if significant tumour growth or increasing catecholamine levels. Here we report the successful use of an observation-only strategy in an infant with unresectable neuroblastoma > 5 cm at diagnosis which initially tripled in size with a 10-fold increase in urinary catecholamines. We highlight the need for consensus evidence-based criteria to define the subgroup where a ‘wait-and-see’ approach is appropriate and criteria to begin active treatment.