To determine whether the type of material used to seal the cochlea after round window cochlear implantation influences delayed hearing loss.Cochlear implants are now prescribed to patients with residual, low-frequency hearing. This hearing-which provides perceptual benefits for the implanted ear-is frequently lost for unknown reasons weeks to months after surgery in a proportion of patients. A post-surgical change in cochlear mechanics, related to the material used to seal the cochlea after round window implantation, may contribute to this loss.An electrode array was implanted in guinea pigs via the round window, which was then sealed with muscle, periosteum, or fibrin glue. Auditory brainstem responses (ABRs) to pure tones (2, 8, 16, 24, and 32 kHz) were recorded before surgery and 1, 4, and 12 weeks after surgery, with subjects then euthanized and their cochleae harvested for histological analysis.Muscle and periosteum, but not fibrin glue, exhibited delayed threshold rises at 2 kHz. Twelve weeks after implantation, 2 kHz threshold shifts differed significantly between muscle (mean, 27.1 dB) and fibrin glue (9.3 dB), but not between these groups and periosteum (19.3 dB). Muscle was sometimes associated with much greater tissue reactions than the other sealants. Most cochleae had injuries to the basilar membrane and/or osseous spiral lamina, regardless of sealant. Hair cell counts did not differ significantly among sealants.Delayed, low-frequency hearing loss was observed when cochleae were sealed with muscle or periosteum, but not when cochleae were sealed with fibrin glue.