The advent of tandem mass spectrometry (MS/MS) around 10 years ago allowed to enlarge consistently the spectrum of metabolic diseases that might be easily and quickly detected. MS/MS was applied to newborn screening in many developed countries, with a wide use, to detect as many as 55 abnormal biochemical conditions (USA), or a restricted one detecting only few diseases (Germany, UK, and Switzerland). Many factors were probably contributing to these very different health organization policies. Although neonatal screening is widely considered extremely useful and efficacious to improve prognosis of many metabolic disorders, the statistically significant demonstration of benefit is quite hard to reach for reasons mainly incidental to the characteristics of these disorders. The expanded newborn screening, in its wide application, includes at present severe diseases presenting in the first days of life, diseases for which treatment is not available, conditions with uncertain significance which are probably not diseases, detection of metabolic disturbances of the mother and all the mildest forms of organic acidurias, urea cycle disorders, fatty acid beta-oxidation defects that may have the possibility to remain asymptomatic for the whole life or may have an acute life-threatening onset of the disease many years later. Which could be the better approach to newborn screening is not clear at present, and probably, it will not be the same for each country. Results of regional screening programs need to be carefully collected and analyzed in future years, with the aim to optimize screening practice in the different countries. Efforts should also be addressed to improve screening programs in the developing countries.