Single-cell proteomics (SCP) has advanced considerably but still is largely limited to homogeneous populations and distant from clinical applicability. We present an SCP workflow for assessing the cellular heterogeneity in saliva and tear fluid. Initially, benchmarks were established using a standard HeLa digestion curve, resulting in more than 5,463 protein groups (PGs) at 50 pg. For single HeLa cells, the workflow was improved to minimize contamination and increase quantitative performance, reaching a maximum of 3,785 PGs per single cell. Following, SCP was benchmarked across heterogenous populations of saliva and tear fluid, collected from 10 healthy individuals. By improving cell isolation, contamination control, and DIA-based search and quantitation, single cells from saliva (n=110) and tear fluid (n=149), with average diameters of 8 and 11 µm, respectively, yielded a maximum of 700 PGs per single cell. Downstream analysis indicated overrepresented protein functions, distinct cluster markers and twenty-three validated therapeutic targets identified from single-cell data. Taken together, this study demonstrates the robustness of our SCP workflow applied to biofluids, driving the discovery of biomarkers and therapeutic targets in complex microenvironments.