Background: Glial fibrillary acidic protein (GFAP), a brain-specific biomarker, shows promise in differentiating intracerebral hemorrhage (ICH) from acute ischemic stroke (IS) and stroke mimics (SM). A novel point-of-care platform measures GFAP in minutes, yet requires centrifugation to obtain plasma. We aim to determine whether participants recruited in an ongoing prospective biomarker study (during working hours) differ from non-recruited patients. Methods: An exploratory analysis of undifferentiated stroke <24h from onset, where plasma GFAP levels (pg/ml) are measured (i-STAT Alinity) at hospital arrival. Clinical characteristics are compared among recruited and non-recruited patients. Results: Among the first 101 patients recruited, mean (±SD) age (70.8±14.5 years), % females (48%), and median (IQR) NIHSS (9(3-20) were similar to the 270 non-recruited patients (70.3±16.3 years, 51% females, NIHSS 7 (3-17), respectively) in the same time period. Median ASPECTS was slightly lower in recruited patients (10(9-10) vs (10(10-10)) (p=0.03). ICH and SM were more common among non-recruited (52% IS/13% ICH/32% SM) compared to recruited patients (67% IS/5% ICH/29% SM, p=0.002), while large-vessel occlusion was more common among those recruited (44% vs 19%, p=0.001). Conclusions: Clinical characteristics do not differ among recruited vs. non-recruited patients in an ongoing biomarker study, yet sampling bias exists regarding underlying stroke condition, with efforts to mitigate this going forward.