Background: Stroke is a time-critical emergency in which rapid diagnosis and treatment are essential to improve outcomes. While Emergency Medical Services (EMS) facilitate structured prehospital stroke care, a proportion of patients bypass EMS and self-present to emergency departments (EDs), potentially causing treatment delays. This study compares process times and outcomes between EMS-transported and self-presenting stroke patients. Methods: We conducted a retrospective analysis of prospectively collected data from 1805 patients with suspected stroke admitted between May 2019 and June 2021 to two hospitals in Germany. Patients were classified as EMS-transported or self-presenting. Process times included door-to-needle time (DNT), symptom onset to first medical contact (FMC), and intravenous thrombolysis (IVT) rates. Data are presented as mean ± standard deviation or median [interquartile range] for continuous variables and as number (%) for categorical variables. Results: A total of 1940 patients with suspected stroke were included. For the main analysis comparing EMS transport and self-presentation, 1805 patients (1525 EMS, 280 self-presenting) were evaluated. EMS patients were significantly older than self-presenting patients (73.1 ± 15.1 vs. 65.3 ± 14.9 years, p < 0.001). The median time from symptom onset to first medical contact was 114 [95-132] minutes in EMS patients and 727 [420-1440] minutes in self-presenting patients ( p < 0.001). IVT was administered in 119/1197 (9.9%) EMS patients and 18/254 (7.1%) self-presenting patients ( p = 0.158) among those with documented information on thrombolysis. Door-to-needle time was slightly but significantly longer in self-presenting patients (median 35 [32-54] vs. 30 [28-34] minutes, p < 0.001), while both groups remained well within certification targets. Conclusions: Self-presenting stroke patients experienced markedly longer prehospital delays and lower intravenous thrombolysis rates compared to EMS-transported patients, although the difference in IVT rates was not statistically significant. In-hospital door-to-needle times were comparable between groups. These findings highlight the need for targeted public education to improve stroke symptom recognition and timely EMS activation.