The purpose of this study was to investigate whether multi-stimulus auditory steady-state responses were capable of estimating hearing thresholds in high-risk infants. A retrospective chart review study. Three tertiary referral centers. Infants born between January 2004 and December 2006 who met the criteria for risk factors of congenital hearing loss were enrolled in the study. While under sedation, the multi-stimulus auditory steady-state response was used to determine multi-channel auditory steady-state response thresholds for high-risk infants younger than 13 months. Conditioned play audiometry was then applied to these children at 23-48 months of age to obtain pure tone audiograms. Auditory steady-state response thresholds and pure tone thresholds were then compared. A total of 249 high-risk infants were enrolled in the study. 39 infants were lost during follow-up. The remaining 216 infants completed both examinations. The Pearson correlation coefficients (r) between the ASSR levels and pure tone thresholds were 0.88, 0.94, 0.94 and 0.97 at 500, 1,000, 2,000 and 4,000 Hz, respectively. The strength of the relationship between the auditory steady-state responses and pure tone thresholds increased with more severe degrees of hearing loss and higher frequencies. We conclude that initial multichannel ASSR thresholds measured under sedation are highly correlated with pure tone thresholds obtained 2 or 3 years later. ASSR can be used to predict the frequency-specific hearing thresholds of high-risk infants and can provide information for early hearing intervention.