To assess the diagnostic validity of morphometric examination of the optic disc and retinal nerve fiber layer (RNFL) thickness to detect permanent structural changes after retrobulbar optic neuritis (ON). Evaluation of a diagnostic test. Twenty-five patients with a history of retrobulbar ON and 29 disease-free controls. The optic discs were evaluated by means of confocal scanning laser ophthalmoscopy (Heidelberg Retinal Tomograph [HRT III]), and RNFL thickness by means of scanning laser polarimetry (GDx), and optical coherence tomography (OCT). Vision function was assessed in all subjects by testing visual acuity, contrast sensitivity, color vision, visual field (VF), and visual evoked potentials (VEPs). Statistical comparisons were made between the affected (ON) and unaffected eyes (non-ON) of the patients with ON, and between these eyes and control eyes (Mann-Whitney test and Wilcoxon's test). Receiver operating characteristic (ROC) curves, and sensitivity and specificity in discriminating ON from control eyes, were calculated for the significant parameters. Correlations between the tests were calculated by means of Spearman's correlation coefficient. We compared OCT, GDx, HRT, and visual testing results in ON eyes versus control eyes. All of the visual function test parameters and RNFL thickness (GDx and OCT) were significantly different between the ON eyes and both the non-ON and control eyes (P<0.01), and there were significant differences in some GDx parameters between the non-ON and control eyes. There were no significant differences in the HRT parameters. The ROC curves indicated that the greatest diagnostic validity was associated with the GDx nerve fiber indicator (AUC, 0.92; sensitivity, 0.80; specificity, 0.97 using a cutoff point of 20.5 between ON and non-ON eyes), and OCT temporal thickness (AUC, 0.92; sensitivity, 0.72; specificity, 0.95 using a cutoff point of 51.5 microm). When investigating permanent damage after ON, RNFL thickness is a promising biomarker. The GDx and OCT are reliable, noninvasive, user-friendly devices; both show good diagnostic validity and good correlations with functional tests in discriminating affected from unaffected eyes. Retinal nerve fiber layer thinning in non-ON eyes should be further studied as a possible subclinical indicator of disease.