OBJECTIVE Transtympanic gentamicin therapy has become a popular treatment modality for Meniere's disease, but questions regarding the ideal dose of medicine, the best administration paradigm, and the safest treatment end-point remain unanswered. The goal of this study is to examine the inner ear kinetics of transtympanic gentamicin and compare this with the kinetics of sustained-release delivery in a basic science model. In addition, we plan to examine the relationship of these kinetics curves to the effect of the two treatment modalities on inner ear function and morphology. It is hoped that this analysis will help clinicians to better apply local medical therapy to the ear. METHODS The study is a basic science project designed to examine perilymph gentamicin concentrations, hearing results, and inner ear morphology in an animal model. METHODS Gentamicin was applied to the right ear of chinchillas either through a transtympanic approach or in a sustained-release device. The left ear remained untreated as an internal control. At set time points the animals' hearing and balance function was studied and the perilymph was harvested, after which the animal was killed and preserved for histological evaluation. Kinetics curves were constructed for each of the two treatment paradigms and compared with histological and functional outcomes. RESULTS The two groups yielded dramatically different kinetics curves. The transtympanic curve had a high peak level at 24 hours with rapid fall-off and almost total elimination by 48 hours, whereas the sustained-release curve was characterized by a long, flat plateau phase with a peak that was approximately one-third that of the transtympanic curve. In addition, the variability seen in perilymph concentrations was significantly higher in the transtympanic group than in the sustained-release group. Immunohistochemical analysis using antibodies against cleaved caspase-3 and cleaved caspase-7 demonstrated early damage in the spiral ganglion of both groups, before any obvious morphological change in the hair cells. The staining was significantly more dense in animals with transtympanic delivery. Cochlear and vestibular hair cell damage was seen at late time points in animals from both groups. Hearing loss (HL) progressed in an orderly fashion in the sustained-release group of animals, with no HL seen in the early time points and universal significant threshold shifts present by 72 hours. In the transtympanic group, the HL was more variable, with significant threshold shifts occurring as early as 4 hours after treatment, but with some animals demonstrating preserved hearing at the 72-hour time point. All animals demonstrated profound HL at the 6-day time point. CONCLUSIONS There is a significant difference in the shape and variability of the perilymph kinetics curve when comparing sustained-release delivery to transtympanic delivery of gentamicin. High early peak levels of gentamicin seen with transtympanic therapy may have a profound effect on the spiral ganglion and produce early HL before obvious hair cell damage. Sustained delivery of gentamicin produces universal HL at 72 hours. The reliability of sustained-release delivery to the ear reduces functional and morphological variations between animals.