Epidemiological and clinical features of patients with in-hospital stroke are not well described in the literature. Our aim was to compare clinical characteristics and outcomes of patients who experienced a stroke during hospital admission with those who were admitted by the emergency department (ED) Retrospective analysis of prospectively collected data from medical records. Clinical and demographic findings were analyzed. Modified Rankin Scale (mRs) was performed at discharge. We included patients who were admitted to a tertiary hospital with any disease and suffered an ischemic stroke (IS), intracerebral hemorrhage (ICH) or subarachnoid hemorrhage (SAH) during admission and those that were admitted by the ED from Feb-09 to Jun-12. We evaluated 501 patients. A total of 11.1% had in-hospital stroke (IHS). In patients with IHS, IS was more frequent 75,8%, followed by ICH 16.0% and SAH 8.9%. Patients who had IHS were similar to those who were admitted by the ED in age, type of stroke, gender and in the use of CT scan for diagnosis. Stroke was more severe in patients with IHS, NIHSS [7.0 (1-9) x 3.0 (3-15), p=0.007]. The median time between admission and stroke was 2.5 days [0,25-6]. When symptoms onset were noticed, 37.7% of the patients were on a regular ward, 32.1% in the ICU, 21.1% in a step down unit (21.1%) and 17.8% in the OR or interventional suite. A total of 39.3% of the patients were admitted for elective surgery, 23.2% for treatment of an infectious, 19.6% for cardiac diseases, 3.6% for cancer treatment and 10.7% for neurological diseases. Patients with IHS and those admitted by the ED were similar in the prevalence of stroke risk factors and in the frequency of complications rates. Thrombolysis rate was similar between the two groups. The mean length of stay was higher in IHS group [20 days 9-47 x 8 days 4-16, p<0.01]. Functional outcome measured by the mRs ≤ 2 at discharge was worse in patients with IHS [37.5% x 68.8%, p<0.005]. Patients with IHS had a higher chance of dying (OR3.52 [95% CI 1,85-6,69]). In conclusion, patients with IHS had a worse prognosis and higher in-hospital mortality when compared to patients admitted because of an incident stroke. This is probably explained by more severe strokes in the IHS group in association with the presence of previous comorbidities