HYPOTHESIS:: The aims of the study were to evaluate tympanometry with regard to age and classification system using two probe-tone frequencies and to provide clinical guidelines. METHODS:: Six subject groups were included in the evaluation: (1) neonatal intensive care unit babies, (2) children younger than 3 months, (3) children 3 to 6 months old, (4) children 6 to 9 months old, (5) children 9 to 32 months old, and (6) adults. Hearing of all subjects was screened by means of auditory brainstem responses, transient-evoked otoacoustic emissions, or behavioral audiometry. Tympanograms, recorded with probe-tone frequencies of 226 and 1000 Hz, were classified according to shape and middle ear pressure. Additionally, 1000-Hz tympanograms were classified based on the Vanhuyse model of tympanometric shapes. Furthermore, tympanometric parameters equivalent ear canal volume, admittance value at +200 daPa, middle ear admittance, tympanometric peak pressure, and tympanometric width were calculated for each tympanogram. RESULTS:: For clinical purposes, the visual admittance classification system was more suitable than the Vanhuyse model. Furthermore, in children younger than the age of 3 months, 1000-Hz tympanometry was easier to interpret and more reliable than 226-Hz tympanometry. From the age of 9 months, 226-Hz tympanometry was more appropriate. In children between 3 and 9 months, the reliability of tympanometry was independent of probe-tone frequency. A two-stage evaluation with a 1000- to 226-Hz tympanometry sequence was preferred because this reduced the total number of tests. CONCLUSION:: The current study provides normative data and age-related guidelines for the use of tympanometry in clinical practice. These results have led to a successful implementation of 1000-Hz tympanometry in neonatal hearing assessment.