A detailed comparison of 2f(1)-f(2) distortion product otoacoustic emission (DPOAE) level shifts (LS) and auditory brainstem response (ABR) threshold shifts with noise-induced histopathology was conducted in chinchillas. DPOAE levels (i.e., L(1) and L(2)) at f(1) and f(2), respectively, ranged from 55-75 dB sound pressure level (SPL), with f(2)/f(1)=1.23, 6 points/octave, f(2)=0.41-20 kHz, and ABR thresholds at 0.5-20 kHz, 2 points/octave, were determined pre-exposure. The exposure was a 108 dB SPL octave band of noise centered at 4 kHz (1-1.75 h, n=6) or 80-86 dB SPL (24 h, n=5). DPOAE LSs (magnitude pre- minus post-exposure) and ABR threshold shifts (TS) were determined at 0 days and up to 28 days post-exposure. The cochleae were fixed, embedded in plastic and dissected into flat preparations. The length of the organ of Corti (OC) was measured; missing inner (IHC) and outer (OHC) hair cells counted; stereocilia damage rated; and regions of OC and nerve-fiber loss determined. Cytocochleograms were made showing functional loss and structural damage with the LS and TS overlaid. Some unexpected results were obtained. First, the best correlation of LS with histopathology required plotting the DPOAE data at f(1) with respect to the chinchilla-place map. The best correlation of TS was with IHC and nerve-fiber loss. Second, wide regions of up to 10% scattered OHC loss in the apical half of the OC showed little or no LS. Third, with the 108 dB SPL noise, there was 20-40 dB of recovery for DPOAEs at mid-high frequencies (3-10 kHz) in eight of 12 cochleae where there was 70-100% OHC loss in the basal half of the OC. The largest recovery at mid-high frequencies occurred in regions where the OC was entirely missing. Fourth, with the 80-86 dB SPL noise, there was no LS at small focal lesions (100% loss of OHCs over 0.4 mm) when the frequency place of either f(1) or f(2) was within the lesion but not both. There was no correlation of LS with OHC stereocilia loss, fusion or disarray. These results suggest that, after noise exposure, DPOAEs at mid-high frequencies can originate from or be augmented by generators located at someplace other than the frequency place of f(2), possibly the basal 20% of the OC when this region is intact. Also, noise-induced DPOAE LSs seemed to reflect differing mechanisms for temporary and permanent hearing loss.