Both basic and clinical electrophysiological investigations have established that the oscillatory potentials (OP) and pattern electroretinogram (PERG) appear to originate from retinal sites that are in proximity. The OPs, subcomponents of the flash ERG, have been shown to reflect disturbances in retinal circulation, and OP amplitude attenuation or loss may be a distinctive feature of diabetic retinopathy. The PERG has been shown to be abnormal in diseases of the optic nerve and ganglion cell body. Thus its relative sensitivity for detection of electroretinal abnormalities in diabetic retinopathy is in question. This study assessed the sensitivity of ERG and OP measures in their detection of abnormalities of electroretinal function in diabetic patients referred to our laboratory. Thirty-five adult Type I patients were studied: 21 with background retinopathy (BR group), 14 with no evidence of background retinopathy (No BR group), and 25 normal control subjects. Monocular OPs were recorded to full-field ganzfeld stimulation at four stimulus intensities. PERGs were obtained from checkerboard pattern reversal stimulation (check-size = 30' arc). Peak-to-peak amplitude and peak implicit time measures of PERGs and OPs were obtained. Subsequent multivariate analysis demonstrated significant differences between normals and diabetic patients, including diabetics with no clinical evidence of retinopathy. In addition, the OP and PERG implicit times appear to be unaffected while OP and PERG amplitudes were diminished in patients with background retinopathy. Only OP amplitudes were found to be significantly diminished in diabetic patients with no photographic evidence of background retinopathy. The PERGs were normal in these patients. Overall, the OP amplitude measures were more sensitive than PERG measures in detecting abnormalities in patients with no retinal photographic evidence of background retinopathy.