OBJECTIVE: Pemoline, a CNS stimulant, is approved for Attention Deficit and Hyperactivity Disorder (ADHD). In practice, it's additionally used to treat narcolepsy and MS-related fatigue. Because pemoline was associated with acute liver failure, FDA relabeled its use as 'second-line' therapy (SLT) for ADHD in 1996. This retrospective observational study examines longitudinal prescription records suggestive of first or second-line use to describe apparent compliance with relabeling and factors associated with second-line use. METHODS: Prescription claims from a continuously-enrolled population (September 1, 2000 - September 30, 2002) from AdvancePCS pharmacy benefit manager (PBM) were evaluated for patients with 1 or more pemoline claims. Patients were categorized using pemoline as SLT or 'first-line' therapy (FLT) depending on presence/absence of other CNS stimulant prescriptions 90 days prior to the first pemoline claim. Chi-square compared SLT and FLT patients with regard to age, gender and prescribing physician specialty. RESULTS: Of 50M covered lives, 2320 had pemoline claims: 51% male, 34% age 0-19 years. 22% used pemoline as SLT (95% CI: 19%-24%). Children < 20 years were more likely to use pemoline as SLT (40%) than adults > 19 years (14%) (p <0.001). Across all ages, males were more likely to use pemoline as SLT (27%) than females (17%) (p <0.001). Physician specialty was associated with SLT (p <0.001): pediatrics (43%), psychiatry (32%), primary care (22%), and neurology (5%). CONCLUSIONS: Most patients in a large PBM appear to use pemoline without antecedent prescriptions for other CNS stimulants, a prescribing pattern inconsistent with FDA-approved SLT labeling. Although ADHD summertime "drug holidays" > 90 days may have inflated estimates of FLT that was actually SLT, children had higher levels of SLT than adults. FDA's labeling action may have affected approved ADHD usage (predominantly male pediatric patients) more than off-label narcolepsy/MS usage (older neurology patients). These results in a large patient population are consistent with earlier, smaller studies.