Sweat-based diagnostics promise to improve Precision Health, where individual health outcomes can be improved through the noninvasive monitoring of specific physiologically relevant chemical signatures that can predict disease risk and onset. Previously, sweat-based diagnostics have only been clinically employed for measurement of sweat chloride levels in cystic fibrosis, due to challenges in reliably measuring a broad array of analytes and understanding which specific ones robustly predict disease.  Realization of sweat-based diagnostics will require understanding the relation between sweat analyte levels and health status. Recent examples have highlighted several possibilities in linking sweat analytes with health status, including tracking sweat uric acid levels for gout, cortisol and cytokine IL-6 levels for stress, lactate levels for physical fatigue and tissue oxygenation status. These approaches share limitations that motivate our work. Notably, these examples along with current sweat chloride clinical measurements in cystic fibrosis are restricted to tracking a couple of specific analytes and require knowing beforehand which analytes are related to disease. In contrast, as a proof of concept, this work applies a simple and rapid unbiased methodology to collect human perspiration samples based on swabbing a glass slide across an individual’s forehead, mass spectrometry methods for detecting a broad set of metabolites in samples, and machine learning approaches for identifying the key metabolic signatures related to end stage renal disease. End stage renal disease, the final stage of chronic kidney disease, is characterized by cessation of kidney function and profound metabolic disturbances, particularly metabolic acidosis, serum lipid abnormalities, changes to glucose homeostasis, and proteinuria. Because routine clinical assessment of chronic kidney disease status using the proteinuria and glomerular filtration rate (eGFR) are known to misdiagnose patients , identifying robust metabolite changes, which are known to be associated with chronic kidney disease progression, has great potential for improving diagnosis and managing kidney disease by providing novel biomarkers.