In a prospective study a population of 32 281 newborn infants in Sweden was screened for signs of respiratory disease. All affected infants (2.9%) were investigated systematically and classified. Incidence and case fatality rates were calculated and related to gestational age, birth weight, sex and postnatal asphyxia, all of which strongly influenced one or both rates. The diagnostic system used was found suitable for future epidemiological research.