This review will summarize the rationale for pursuing investigations into the use of retinoids for pediatric patients with cancer, describe the phase I results of all-trans-retinoic acid (ATRA) in children, and discuss the results of a series of preclinical and clinical pharmacokinetic studies of ATRA. The prognosis for children with advanced stage neuroblastoma, the most common extracranial solid tumor of childhood, has remained poor despite significant increases in the intensity of multi-modality therapy. Observations that neuroblastoma has the potential in vivo to differentiate into the more mature neuronal phenotype of a ganglioneuroma or to spontaneously regress, combined with the ability of ATRA to induce differentiation of neuroblastoma cell lines in vitro, suggests that neuroblastoma may be a prime candidate for a retinoid-based approach to differentiation therapy. We previously performed a standard pediatric phase I trial of ATRA and defined the maximum tolerated dose (MTD) in children to be 60 mg/m2/day, significantly lower than the MTD in adult patients. Pharmacokinetic results revealed that the plasma half-life of ATRA was short (45 min) relative to 13-cis-RA (12-24 h), and that plasma drug exposure decreased significantly by day 28 of daily drug administration. Preclinical studies using an i.v. formulation of ATRA in a Rhesus monkey pharmacokinetic model then demonstrated that ATRA is eliminated by a capacity-limited (saturable) process. This elimination process was rapidly induced within the first week of daily i.v. ATRA administration, and suggested that an intermittent schedule of drug administration might allow for down-regulation of the elimination process. These pre-clinical studies formed the basis for investigating whether an intermittent schedule of ATRA administration would allow for repeated periods of relatively higher plasma drug concentrations. Preliminary results of two clinical trials using intermittent schedules of administration suggest that this approach may result in significantly higher plasma drug exposure over time. Plans to study the role of intermittent schedules of ATRA administration in pediatric phase II trials in patients with neuroblastoma are underway.