BACKGROUND: Obtaining lower airway specimens is important for guiding therapy in chronic lung infection but is difficult in young children unable to expectorate. While culture-based studies have assessed the diagnostic accuracy of nasopharyngeal or oropharyngeal specimens for identifying lower airway infection, none have used both together. We compared respiratory bacterial pathogens cultured from nasopharyngeal and oropharyngeal swabs with bronchoalveolar lavage (BAL) cultures as the "gold standard" to better inform the diagnosis of lower airway infection in children with chronic wet cough. METHODS: Nasopharyngeal and oropharyngeal swabs and BAL fluid specimens were collected concurrently from consecutive children undergoing flexible bronchoscopy for chronic cough and cultured for bacterial pathogens. RESULTS: In cultures from 309 children (median age, 2.3 years) with chronic endobronchial suppuration, all main pathogens detected (Haemophilus influenzae, Streptococcus pneumoniae, and Moraxella catarrhalis) were more prevalent in nasopharyngeal than oropharyngeal swabs (37%, 34%, and 23% vs 21%, 6.2%, and 3.2%, respectively). Positive and negative predictive values for lower airway infection by any of these three pathogens were 63% (95% confidence interval [95% CI] 55, 70) and 85% (95% CI, 78, 91) for nasopharyngeal swabs, 65% (95% CI, 54, 75), and 66% (95% CI, 59, 72) for oropharyngeal swabs, and 61% (95% CI, 54,68), and 88% (95% CI, 81, 93) for both swabs, respectively. CONCLUSIONS: Neither nasopharyngeal nor oropharyngeal swabs, alone or in combination, reliably predicted lower airway infection in children with chronic wet cough. Although upper airway specimens may be useful for bacterial carriage studies and monitoring antimicrobial resistance, their clinical utility in pediatric chronic lung disorders of endobronchial suppuration is limited.