Methylphenidate (MPH) has been shown to be an effective and cost-effective treatment for children and adolescents with attention-deficit/hyperactivity disorder (ADHD). Given their short duration of action, MPH immediate-release (MPH-IR) formulations require multiple daily dosing. Studies have reported noncompliance rates of 20-65%. OBJECTIVE: To evaluate, from the perspective of the UK NHS, the cost-effectiveness of MPH-OROS, a novel once-a-day formulation. METHODS: A meta-analysis was performed to synthesize data on clinical efficacy from three randomized clinical trials, comparing MPH-OROS o.a.d., MPH-IR t.i.d., and placebo (Caro 2003). Results were combined with unit cost data (BNF 2003, PSSRU 2003), resource utilization estimates (NHS perspective), and assumptions on treatment compliance (systematic review, Claxton et al. 2001). Data were used to populate a decision tree model adapting and extending the CCOHTA analysis (1998) of ADHD therapies. RESULTS: MPH-OROS and MPH-IR were significantly more efficacious than placebo, in both community teacher and parent ratings of inattention/overactivity (IOWA Conners I/O scale; primary trial endpoint). For teacher ratings, standardized mean differences (SMD, random effects model) compared to placebo were 1.32 (1.09-1.55, 95% CI, for MPH-OROS) and 1.19 (1.00-1.38 for MPH-IR); effect sizes reported by parents were generally higher and better for MPH-OROS compared to MPH-IR. Assuming 79% compliance with MPH-OROS o.a.d. and 65% with MPH-IR t.i.d. over one year, the incremental cost of MPH-OROS per SMD (teacher ratings) was £1345 (for MPH-IR: £1120); for parent ratings, MPH-OROS exhibited extended dominance over MPH-IR. Comprehensive sensitivity analyses were performed. For MPH-IR compliance rates below 57%, MPH-OROS dominated (in an extended sense) also in teacher ratings. CONCLUSION: These data indicate that MPH-OROS may be more effective than MPH-IR t.i.d. in daily practice. They suggest an acceptable incremental cost-effectiveness ratio of MPH-OROS, with extended dominance over MPH-IR under a broad range of assumptions. Real world data will have to confirm these estimates.