Abstract : Objectives : To analyse published cost-of-illness studies that had assessed the cost of prematurity according to gestational age at birth. Methods : A review of the literature was carried out in March 2011 using the following databases: Medline, ScienceDirect, The Cochrane Library, Econlit and Business Source Premier, and a French Public-Health database. Key-word sequences related to ‘prematurity’ and ‘costs’ were considered. Studies that assessed costs according to the gestational age (GA) at the premature birth (<37 weeks of gestation) in industrialized countries and during the last two decades were included. Variations in the reported costs were analysed using a check-list, which allowed the studies to be described according to several methodological and contextual criteria. Results : A total of 18 studies published since 1990 were included. According to these studies, costs were assessed for different follow-up periods (short, medium or long-term), and for different degrees of prematurity (extreme, early, moderate and late). Results showed that whatever the follow-up period, costs correlated inversely with GA. They also showed considerable variability in costs within the same GA group. Differences between studies could be explained by the choices made, concerning i/the study populations, ii/contextual information, iii/and various economic criteria. Despite these variations, a global trend of costs was estimated in the short-term period using mean costs from four American studies that presented similar methodologies. Costs stand at over US\ 100,000 for extreme prematurity, between US\ 40,000 and US\ 100,000 for early prematurity, between US\ 10,000 and US\ 30,000 for moderate prematurity and below US\ 4500 for late prematurity. Conclusion : This review underlined not only the clear inverse relationship between costs and GA at birth, but also the difficulty to transfer the results to the French context. It suggests that studies specific to the French health system need to be carried out.