Abstract The significance of fluid metabolism among the patients with cerebral infarction has barely mentioned in the literature despite the several reports suggesting the potential risk of reduced hydration status for the development of cerebral infarction. The aim of the this study is to explore the validity of the presumable relationship between hydration status and cerebral infarction. Ninety-seven patients with cerebral infarction from April 1, 2008 to March 31, 2009 were retrospectively investigated, and their hydration status were evaluated by using several clinical parameters such as a blood urea nitrogen to serum creatinine (BUN/Cr) ratio of >25 and plasma osmolality. Subjects with active infection, congestive heart failure, hepatic failure, gastrointestinal bleeding, or a malignancy were excluded since these conditions should modulate the absolute value of BUN/Cr ratio without a change in hydration status. Twenty-eight patients (29%) were considered as having reduced hydration status. The BUN/Cr ratio decreased significantly after the initiation of medical support (median 21.3; IR: 18.1–24.6), including oral or parenteral fluid supplementation, in comparison to the values at the time of patient admission (median 30.0; IR: 26.8–40.7; p < 0.0001). Similar decreases were also observed in the hematocrit, hemoglobin, and plasma osmolality. The group considered to have reduced hydration status had a significantly higher prevalence of cardioembolic stroke than the other subjects. The hydration status may be a contributing factor to subtypes of cerebral infarction. Whether our findings are also the case with overall patients with cerebral infarction should be evaluated in greater detail.