The effects of type 2 diabetes on evoked otoacoustic emissions (e-OAEs) elicited by clicks in subjects with normal hearing and the involvement of the central (CNS) and peripheral nervous system and acute hyperglycemia were investigated. In study 1, 110 type 2 diabetic patients and 106 control subjects matched for age and gender were investigated by e-OAEs. Central and peripheral neuropathy were evaluated respectively by auditory brainstem responses (ABRs) and according to San Antonio Consensus Conference criteria. In study 2, 10 healthy and 10 type 2 diabetic men matched for age, all with normal e-OAEs, underwent a 5-hour hyperglycemic clamp study. e-OAE tests were performed before and during the hyperglycemic clamp. In study 1, e-OAEs were impaired in 51.8% (57 of 110) of the diabetic subjects, in comparison to 4.7% (five of 106) of the control group (P < .0001). Diabetics with impaired e-OAEs (e-OAEs-), in comparison to those with normal e-OAEs (e-OAEs+), were older (51.0+/-5.8 v 45.1+/-6.0 years, P < .001), had diabetes longer (11.5+/-4.4 v 7.0+/-3.9 years, P < .001), achieved poorer metabolic control as judged by hemoglobin A1c ([HbA1c] 6.9%+/-0.4% v 6.5%+/-0.3%, P < .001), and had more peripheral neuropathy (46% v 23%, P < .02). No difference was observed between e-OAEs- and e-OAEs+ subjects for retinopathy or nephropathy. Nevertheless, when the duration of diabetes was corrected by multiple regression analysis, the correlation between sensorineural damage and peripheral neuropathy lost significance (P = .12). Diabetic groups (e-OAEs+ and e-OAEs-) showed greater latency in waves I, III, and V and greater interwave latency for waves I to V than the control group, but there was no significant difference in ABRs between e-OAEs+ and e-OAEs- subjects. In study 2, there were no significant changes in e-OAE intensities compared with basal values during the entire hyperglycemic clamp in either type 2 diabetic or control subjects. No difference was observed between the two groups at each time of the clamp. Thus, type 2 diabetic subjects show a higher rate of compromised e-OAEs than healthy individuals. The e-OAE dysfunction does not associate with either an injury to the auditory nervous pathway or diabetic microvasculopathy. The apparent interference of peripheral neuropathy in e-OAEs loses significance when corrected for the duration of diabetes.