The aim of this epidemiological study was to utilise a cross-sectional as well as a longitudinal approach to examine sleep habits and how they develop in young people in Iceland. The 668 subjects (1-20 years) who responded to a postal survey in 1985 were followed up 5 and 10 years later. The majority of the variance in bedtime and sleep duration was explained by age, but also to a considerable degree by other factors such as residence, season, and year of survey or interaction of these factors. Natural phenomena, such as the diminution of total sleep duration in the first years of life and the tendency for longer sleep on weekends compared to weekdays were confirmed. The lengthening of sleep on weekends was first significant at the age of 9 and was greater among adolescents than young adults. The incidence of daytime sleepiness increased in adolescence, as did napping, at which time their nocturnal sleep time significantly decreased. Over a period of 10 years, a significant shift to earlier wake-up times occurred in children up to 15 years of age, which resulted in a shortened total sleep time. The idea that individual sleep duration is an inherent parameter is supported by the high positive correlation of total sleep time across a 10-year period (r=.73). The present data confirm that Icelandic adolescents (aged 11, 13, and 15) have delayed bedtimes and shorter nocturnal sleep compared to European peers.