Introduction:. Few studies have tested the hypothesis of whether the beneficial effect of Stroke Units (SUs) can be reproduced in routine clinical practice and whether SU are also superior to neurological wards [NWs]. We aimed to compare the outcomes of patients of a newly implemented SU to the outcomes of patients hospitalized in a NW. Methods:. We made a before-after comparison of 352 SUs and 352 NWs patients after adjusting for case-mixes by the multivariate method. Subgroup analyses were also performed to evaluate which patient groups benefit the most. In-hospital case-fatality, proportion of independent patients at discharge, length of hospital stay (LOHS), medical complication rate were the main outcome measures. Results:. Adjusted in-hospital case fatality was significantly reduced in the SUs (OR: 0.44, 95 % CI: 0.26-0.76; p = 0.003). The proportion of independent patients at discharge and patients having medical complications was not different. Length of hospital stay was shorter in SU patients (13.76 days vs. 16.72 days, p = 0.003). Treatment in the SUs decreased case fatality in many subgroups [men, elderly, early admitted, severe stroke, co-morbidity present and ischemic stroke groups]. Discussion:. The results of randomized trials in favor of SUs can be reproduced in routine clinical practice. The benefit of SU care seems to be more apparent with advancing age and increasing stroke severity. Stroke Unit seems to be a better alternative to an experienced NW.