ABSTRACT Aims To determine the utility of corneal confocal microscopy and tear neuromediator analysis in the diagnosis of diabetic peripheral neuropathy (DPN) as a result of type 1 and type 2 diabetes. Methods Seventy individuals with either type 1 diabetes or type 2 diabetes (T1D/T2D) underwent corneal confocal microscopy to assess the corneal nerve morphology. The concentration of substance P and calcitonin gene-related peptide (CGRP) in tears was measured by enzyme-linked immunosorbent assay. Motor excitability studies were conducted on the median nerve to assess axonal ion channel function. Based on total neuropathy score (TNS), participants were stratified into DPN (DPN+ve; TNS ≥ 2; T1D, n = 19; T2D, n = 16) and without DPN (DPN–ve; TNS ≤ 1; T1D, n = 19; T2D, n = 16). Areas under the receiver operating characteristic curves (AUCs) were calculated to obtain specificity and sensitivity of the measures to diagnose DPN. Results In T1D, the concentration of substance P and confocal microscopy measures were significantly reduced (P < .010) in DPN+ve. Also, for the nerve excitability measures, mean peak response, percentage of threshold electrotonus at peak and after 90-100 ms, superexcitability and subexcitability were significantly reduced (P < .050) in DPN+ve. In T2D, except for inferior whorl length (P = .190), all other corneal confocal microscopy measures were significantly reduced (P < .010) in DPN+ve, but there was no difference in substance P concentration. For the diagnosis of DPN in T1D, the AUC for inferior whorl length (0.910), mean peak response (0.800) and concentration of substance P (0.770) were high and in T2D, the AUC for corneal nerve fiber length (0.809) and nerve fractal dimension (0.777) were high. Conclusion Corneal confocal microscopy parameters provide a better diagnostic ability to detect DPN in T1D and T2D than nerve excitability measures or concentrations of tear neuromediators. The concentration of substance P could also be useful in diagnosing DPN but for T1D only.