BACKGROUND: To address the major methodological issues of reverse causation and selection bias in epidemiologic studies of antibiotic use in early life and the development of asthma, we undertook a cohort study of this association in a complete population of children. METHODS: Using the health-care and prescription databases of Manitoba, Canada, this longitudinal study assessed the association between antibiotic prescription use during the first year of life and asthma at age 7 years in a 1995 birth cohort of 13,116 children. RESULTS: Independent of well-known asthma risk factors, asthma was significantly more likely to develop in children who had received antibiotics in the first year of life at age 7 years. The association with asthma was observed for antibiotic use in non-respiratory tract infections (adjusted odds ratio [OR], 1.86; 95% confidence interval [CI], 1.02 to 3.37). The risk of asthma was highest in children receiving more than four courses of antibiotics (adjusted OR, 1.46; 95% CI, 1.14 to 1.88), especially among rural children, and in the absence of maternal asthma or a dog in the birth year. Broad-spectrum (BS) cephalosporin use was more common in these subpopulations of children. CONCLUSIONS: Antibiotic use in early life was associated with the development of childhood asthma, a risk that may be reduced by avoiding the use of BS cephalosporins.