The importance of population-based long-term follow-up studies of geographically determined cohorts to evaluate the effectiveness, efficiency and availability of a regionalized perinatal-neonatal care programme is demonstrated by the Victorian Infant Collaborative Study Group. The survival and quality of survival of consecutively born extremely-low-birthweight infants below 1000 g or extremely preterm infants below 28 weeks' gestation in the state of Victoria were assessed up to 14 years of age over four distinctive eras: 1979-1989, 1985-1987, 1991-1992 and 1997. Both survival and quality-adjusted survival rates rose progressively in all birth weight and gestation subgroups, associated with progressively more such infants being born in level III perinatal centres. Cost-effectiveness and cost-utility ratios remained stable overall, with efficiency gains in the smaller infants over time. Regionalized long-term follow-up provides unique information that is not available from institution-based studies, which is vital to the regional organization of perinatal-neonatal care.