Background: Plasma protein biomarkers measured in the hyperacute setting that are associated with stroke diagnosis or with clinical neurological severity, can be diagnostically and prognostically meaningful in patient care. We hypothesized that high-throughput proteomics can identify panels of plasma proteins that can differentiate stroke mimics from ischemic and hemorrhagic stroke, and are associated with neurological severity. Methods: From a clinically-characterized biorepository of 320 plasma samples collected at time of patient presentation to the Emergency Room at Grady Memorial Hospital (Atlanta), 40 samples (10 stroke mimics, 10 intracerebral hemorrhage [ICH], 20 acute ischemic stroke [AIS]) were included (Fig A). 7,000 proteins were measured per sample using aptamer-based proteomics (SomaLogic). A combination of differential-expression and variance partitioning analyses (VPA) identified proteins associated with diagnosis (mimic vs ICH vs AIS), neurological severity (NIHSS), age and sex. Results: Cohort characteristics are shown (Fig B). VPA identified top proteins that accounted for majority of variance in several traits (Fig C). One-way ANOVA identified >400 proteins showing group-wise differences, of which a panel of 56 proteins collectively distinguished mimics from both AIS and ICH, ICH from both mimics and AIS, as well as distinguished between AIS based on NIHSS (Fig D). The pattern of expression of these 56 plasma proteins across diagnostic groups is shown in Fig E. We also identified plasma biomarkers associated with increasing age, indicative of age-related vascular dysfunction. Conclusions: This pilot study of AIS, ICH and stroke mimics demonstrates the applicability of aptamer-based proteomics as a promising approach for plasma protein biomarker discovery in stroke. We identified acute plasma protein biomarkers of neurological severity, diagnostic group and age-related vascular dysfunction that need further validation.