OBJECTIVE To establish the prevalence of, and risk factors associated with, diabetic retinopathy in an Australian adolescent diabetes clinic population. METHODS A prospective longitudinal study; baseline findings. METHODS Two hundred and fifty-five patients with Type 1 (insulin-dependent) diabetes mellitus assessed by our service were studied. Entry criteria were: age 11.0-19.9 years; diabetes duration of at least two years; and gradable fundus photographs of at least one eye. METHODS The presence and severity of retinopathy, as assessed by the grading of stereoscopic fundus photographs. Possible risk factors assessed were age, sex, diabetes duration, pubertal stage, blood pressure, glycaemic control and total cholesterol level. RESULTS The prevalence of retinopathy was 42%; all of those affected had mild background retinopathy. Highly significant associations were found with glycaemic control and both total and prepubertal duration of diabetes. No associations were found with age, sex, pubertal stage, blood pressure or total cholesterol level. CONCLUSIONS The high prevalence of early diabetic retinopathy in this group of Australian adolescents is comparable to recent reports from other centres. The significant associations with glycaemic control and duration of diabetes provide further strong evidence for the benefit of optimal glycaemic control during adolescence. Our finding that the prepubertal years of diabetes contribute to the development of retinopathy suggests that glycaemic control before puberty should also be optimised. The planned follow-up of this cohort will establish the risk of progression to vision-threatening retinopathy and allow risk factors to be further evaluated.