To investigate whether the rising incidence of Type 1 diabetes in children is evident in young adults and determine whether age at onset has decreased over time. Two geographically defined datasets from the population-based Yorkshire Diabetes Register were analysed: (i) 2718 children diagnosed under 15 years with Type 1 diabetes from 1978 to 2000 in Yorkshire; (ii) 631 young adults (15-29 years) diagnosed from 1991 to 1999 in West Yorkshire. Log-linear regression and age-period-cohort modelling evaluated changes in incidence over time and age at onset. Incidence rose steadily for 0-14-year-olds in Yorkshire with an average annual increase of 2.9%[95% confidence interval (CI) 2.0, 3.8]. In West Yorkshire between 1991 and 1999, the time trends for 0-14 and 15-29-year-olds were significantly different (P = 0.014). Stable rates in 15-29-year-olds contrasted with an average annual increase of 5.9% (95% CI 2.7, 9.2) for 0-14-year-olds. The mean age at onset fell from 9.2 to 8.4 years for 0-14-year-olds and from 16.0 to 14.6 years for 0-29-year-olds. Age-period-cohort modelling showed a statistically significant (P < 0.001) increased risk of developing diabetes was associated with decreasing age for those diagnosed more recently. A steady and continuing rise in the incidence of Type 1 diabetes over time is observed for children but not for young adults. In parallel, the age at onset is gradually decreasing and more recent birth cohorts are at increased risk. This overall pattern is consistent with the influence of an environmental agent that is gradually affecting children at younger and younger ages.