Biomarkers in tear fluid have attracted much interest in daily healthcare sensing and monitoring. Recently, surface-enhanced Raman scattering (SERS) has enabled highly sensitive label-free detection of small molecules. However, a highly stable straightforward tear assay with superior sensitivity is still under development in tear collection and analysis. Here we report a plasmonic Schirmer strip for on-demand, rapid, and simple identification of biomarkers in human tears. The diagnostic strip features gold nanoislands directly and evenly formed on the top surface of cellulose fibers, which maintain a hygroscopic nature for an efficient collection of tear production as well as provide plasmonic enhancement in SERS signals for identification of tear molecules. The uric acid in human tears was quantitatively detected at physiological levels (25-150 μM) by using SERS. The experimental results also clearly reveal a strong linear correlation between uric acid level in both human tears and blood for gouty arthritis diagnosis. This functional paper strip enables noninvasive diagnosis of disease-related biomarkers and healthcare monitoring using human tears.