BACKGROUND: The aim of this study was to assess the prevalence of undernutrition in patients in Dutch rehabilitation centres and to measure the diagnostic accuracy of available screening tools. METHODS: This cross-sectional multicentre study was conducted in 11 rehabilitation centres in The Netherlands. Patient's nutritional status was determined by the amount of weight loss during the last 1, 3 and 6 months and body mass index (BMI). Diagnostic accuracy was assessed for 5 screening tools: Short Nutritional Assessment Questionnaire (SNAQ), Short Nutritional Assessment Questionnaire Residential Care (SNAQRC), SNAQ65+, Malnutrition Universal Screening Tool and Mini nutrition Assessment-short form. RESULTS: Twenty-eight percent of the patients were severely undernourished and 10% were moderately undernourished. In the undernourished group, 28% were overweight (BMI 25-30) and 19% were obese (BMI > 30). The SNAQ65+ is the recommended screening tool due to its high diagnostic accuracy (sensitivity 96%, specificity 77%, positive predictive value 62%, negative predictive value 90%) and quick and easy use. The MNA had the worst diagnostic accuracy, with a sensitivity of 44%. CONCLUSION: The prevalence of undernutrition in patients in Dutch rehabilitation centres is high. Almost half of the undernourished patients were overweight or obese. Therefore, it is important not only to screen for undernutrition, but also carefully to assess possible overweight/obesity in every undernourished rehabilitation patient.