OBJECTIVES. To systematically review the literature on inborn errors of metabolism, neonatal screening technology and screening programmes in order to analyse the costs and benefits of introducing screening based on tandem mass-spectrometry (tandem MS) for a wide range of disorders of amino acid and organic acid metabolism in the UK. To evaluate screening for cystic fibrosis, Duchenne muscular dystrophy and other disorders which are tested on an individual basis. HOW THE RESEARCH WAS CONDUCTED. Systematic searches were carried out of the literature on inborn errors of metabolism, neonatal screening programmes, tandem MS-based neonatal screening technology, economic evaluations of neonatal screening programmes and psychological aspects of neonatal screening. Background material on the biology of inherited metabolic disease, the basic philosophy, and the history and current status of the UK screening programme was also collected. Relevant papers in the grey literature and recent publications were identified by hand-searching. Each paper was graded. For each disease an aggregate grade for the state of knowledge in six key areas was awarded. Additional data were prospectively collected on activity and costs in UK neonatal screening laboratories, and expert clinical opinion on current treatment modalities and outcomes. These data were used to construct a decision-analysis model of neonatal screening technologies, comparing tandem MS with the existing phenylketonuria screening methods. This model determined the cost per additional case identified and, for each disease, the additional treatment costs per case, and the cost per life-year saved. All costs and benefits were discounted at 6% per annum. One-way sensitivity analysis was performed showing the effect of varying the discount rate, the incidence rate of each disorder, the number of neonates screened and the cost of tandem MS, on the cost per life-year gained. RESEARCH FINDINGS. The UK screening programmes for phenylketonuria and congenital hypothyroidism have largely achieved the expected objectives and are cost-effective. Current concerns are the difficulty of maintaining adequate coverage, perceived organisational weaknesses, and a lack of overview. For many of the organic acid disorders it was necessary to rely on data obtained from clinically-diagnosed cases. Many of these diseases can be treated very effectively and a sensitive screening test was available for most of the diseases. Except for cystic fibrosis, there have been no randomised controlled trials of the overall effectiveness of neonatal screening. Despite the anxiety generated by the screening process, there is strong parental support for screening. The effects of diagnosis through screening on subsequent reproductive behaviour is less clear. Conflicts exist between current concepts and the traditional principles of screening. The availability of effective treatment is not an absolute prerequisite: early diagnosis is of value to the family concerned and, to the extent that is leads to increased use of prenatal diagnosis, may help to reduce the overall burden of disease. Neonatal screening is also of value in diseases which present early but with non-specific symptoms. Indeed, almost all of the diseases considered could merit neonatal screening. The majority of economic evaluations failed to incorporate the health benefits from screening, and therefore failed to address the value of the information which the screening programmes provided to parents. The marginal cost of changing from present technology to tandem MS would be approximately 0.60 pounds per baby at a workload of 100,000 samples a year, and 0.87 pounds at 50,000 samples per year. The ability to screen for a wider range of diseases would lead to the identification of some 20 additional cases per 100,000 infants screened, giving a laboratory cost per additional diagnosis of 3000 pounds at an annual workload of 100,000 babies per year.(ABSTRACT TRUNCATED)