BACKGROUND: Nutrition's impact on stroke rehabilitation outcomes is controversial. Existing studies utilize albumin without correcting for inflammation in nutritional assessments. Here, prealbumin was used and inflammation assessed to determine if nutrition impacts rehabilitation outcomes. OBJECTIVE: Determine the effect of dietary intake on prealbumin level, number of complications, length of stay, and Functional Independence Measure (FIM) efficiency in rehabilitation stroke inpatients. METHODS: Patients had admission and discharge prealbumin and C-reactive protein (CRP) levels drawn; and, weekly protein and calorie counts obtained. Patients were followed for number of complications, length of stay, and FIM efficiency. RESULTS: Mean protein and calorie intake was 57.6 ± 16.2 g/d and 1452.2 ± 435.8 kcal/d, respectively. 77.6% of patients had normal prealbumin on admission with 94.9% on discharge. Prealbumin increased significantly from admission to discharge (22.3 ± 6.2 mg/dL vs. 24.6 mg/dL ± 5.1 mg/dL, P = 0.007). Number of complications and length of stay were predicted by CRP in regression models. Total, motor, and cognitive FIM efficiencies were not universally affected by prealbumin levels, protein intake, or calorie intake. CONCLUSIONS: Nearly all hypoprealbuminemic stroke rehabilitation inpatients correct their levels eating a non-supplemented diet. Number of complications, length of stay, and functional outcomes in this patient are not affected by prealbumin levels, protein intake, or calorie intake.