Little is known about cisplatin ototoxicity in pediatric patients. Measurement of otoacoustic emissions is a rapid, reproducible, objective method of evaluating hearing. We examined whether transient-evoked otoacoustic emissions in pediatric patients exposed to cisplatin in the past correlated with audiographic findings. Twelve patients were entered into the study (mean age at treatment 7.8 years, mean cumulative dose 442.5 mg/mm2, mean 7.1 doses). Hearing at 3000 Hz was preserved in 82.6% of patients. In the higher frequencies significant sensorineural hearing loss was noted: 43.5% at 4 kHz; 81.0% at 6 kHz; and 90.5% at 8 kHz. Transient-evoked otoacoustic emissions were measurable in 11 of 12 patients. Middle ear disease accounted for abnormal otoacoustic emission seen in three patients (1 with effusion, 2 with significant negative middle ear pressure). When the middle ear was normal, a statistically significant correlation was seen between the transient-evoked otoacoustic emissions reproducibility and pure-tone threshold (correlation coefficient = -0.69, p = 0.008). Increased hearing loss was also associated with young age at first dose of cisplatin (p = 0.044), high number of chemotherapy cycles (p = 0.042), and high cumulative dose (p = 0.042).