BACKGROUND: Cough in children is a common reason for medical consultations and affects quality of life. There are little population-based data on the epidemiology of recurrent cough in children and how this varies by age and sex, or between children with and without wheeze. We determined the prevalence of cough throughout childhood, comparing several standardised cough questions. We did this for the entire population and separately for girls and boys, and for children with and without wheeze. METHODS: In a population-based prospective cohort from Leicestershire, UK, we assessed prevalence of cough with repeated questionnaires from early childhood to adolescence. We asked whether the child usually coughed more than other children, with or without colds, had night-time cough or cough triggered by various factors (triggers, related to increased breathing effort, allergic or food triggers). We calculated prevalence from age 1 to 18 years using generalised estimating equations for all children, and for children with and without wheeze. RESULTS: Of 7670 children, 10% (95% CI 10-11%) coughed more than other children, 69% (69-70%) coughed usually with a cold, 34% to 55% age-dependently coughed without colds, and 25% (25-26%) had night-time cough. Prevalence of coughing more than peers, with colds, at night, and triggered by laughter varied little throughout childhood, while cough without colds and cough triggered by exercise, house dust or pollen became more frequent with age. Cough was more common in boys than in girls in the first decade of life, differences got smaller in early teens and reversed after the age of 14 years. All symptoms were more frequent in children with wheeze. CONCLUSIONS: Prevalence of cough in children varies with age, sex and with the questions used to assess it, suggesting that comparisons between studies are only valid for similar questions and age groups.