There is still controversy regarding the relative merits of catecholamine metabolite estimations on 24 h versus untimed urine collections. The former has the advantage of taking into account diurnal variation in the rate of metabolite excretion but has the disadvantages of delaying results and of being affected by errors in collection. In this study percentile values were established for a reference population of 181 children for urinary 4-hydroxy-3-methoxyphenylacetic acid (HVA)/creatinine and 163 children for 4-hydroxy-3-methoxymandelic acid (VMA)/creatinine, using untimed urine collections. Results of similar determinations performed as part of the diagnostic work up of 23 consecutive children subsequently proven to have neural crest tumours showed that all patients had the value of at least one metabolite concentration at or above the highest reference value. In neuroblastoma all patients' VMA/creatinine exceeded the highest reference value and in neural crest tumours overall, this ratio was greater than the highest reference value in 96% of patients. These results are as good as, or better than, previously published results and demonstrate the practical value of using catecholamine metabolite determinations expressed as 'creatinine equivalents' on untimed urine specimens in the diagnosis of neuroblastoma and related tumours in children.