Tear biomarkers have received much attention for non-invasive monitoring of ocular and systemic diseases. Although various approaches have been developed for on-site detection of tear biomarkers, there is still the limited number of simple and reliable sensing platform in the clinical setting. Here, we introduce a plasmonic enzyme-linked immunosorbent assay (ELISA) that utilizes Ag/chitosan plasmonic paper with dual-signal readout to detect immunoglobulin G as a model tear biomarker. The sensing mechanism for a colorimetric and surface-enhanced Raman scattering (SERS) dual-signal readout is attributed to the etching of Ag nanoparticles anchored to the plasmonic paper. This etching induces both a color shift and a reduction in the Raman enhancement of the plasmonic paper. The optimized etching system demonstrates the feasibility of performing simultaneous immunoassay and etching on a plasmonic substrate without requiring an additional etching step. The detection limit was achieved at concentrations approximately 100 times lower than the physiological tear concentration, along with selective detection in simulated tear fluid. Furthermore, the dual-mode detection improves accuracy and efficiency by comparing the results of overlap detection range and enabling pre-screening with colorimetric results prior to SERS analysis. This approach holds promise as a versatile point-of-care sensing platform that can easily and reliably detect tear biomarkers.