Diagnostic indications for flexible bronchoscopy in the initial investigation of children with suspected foreign-body (FB) aspiration have not been evaluated prospectively. We prospectively collected history, clinical, and radiologic findings at prebronchoscopic examination of all children referred for suspected FB aspiration between February 1993 and September 1995. Children with asphyxiating FB aspiration, requiring immediate rigid bronchoscopy, were excluded. If there was clear evidence of FB aspiration from the physical and radiographic findings, rigid bronchoscopy was directly performed. If the evidence was not convincing, children underwent diagnostic flexible bronchoscopy under local anesthesia. If an FB was found, rigid bronchoscopy was always performed for extraction. Eighty-three consecutive children (median age: 24 mo) were included. Among 28 who underwent rigid bronchoscopy first, 23 had an FB. Among the 55 children who underwent flexible bronchoscopy first, 17 had an FB. Predictive signs of a bronchial FB were a radiopaque FB, and associated unilaterally decreased breath sounds and obstructive emphysema (positive predictive value = 0.94). We propose the following management algorithm: Rigid bronchoscopy is performed first in case of asphyxia, a radiopaque FB, or association of unilaterally decreased breath sounds and obstructive emphysema. In any other case, flexible bronchoscopy is performed first for diagnostic purposes. If applied retrospectively to the 83 children in our study, this algorithm would have decreased the negative first rigid bronchoscopy rate to 4%. Flexible bronchoscopy is a safe and cost-saving diagnostic procedure in children with suspected FB aspiration.