Protein-energy malnutrition (PEM) in hospitalized patients is associated with poor outcomes. Prealbumin (PAB) has been shown to be a potentially useful tool in PEM assessment. PAB plasma concentrations are closely related to nutritional state and nutritional intervention. However, PAB is also inversely correlated to C-reactive protein (CRP), making difficult the interpretation of PAB changes during tratment of severely ill patients with inflammation and/or infection. Methods: To assess the confounding role of CRP we investigated the role of two seriate measures of PAB and CRP in monitoring the response to nutritional intervention in 83 hospitalized patients enrolled in a nutritional care program (group 1). Data were also collected from 60 patients with low PAB not enrolled in any nutritional program (group 2). Results: When nutritional needs were not satisfied (nutritional intake/needs ratio <75%) we found no PAB increase, whereas there was a significant PAB increase in patients with adequate intake (mean delta-PAB 4.2 mg/dL). Covariance analysis showed a statistically significant mean PAB increase of 2.38 mg/dL in the adequate intake subset, value controlled for the confounder delta-CRP (p < 0.005). In the inadequate intake subset and in group 2 an improvement in CRP status was not associated with a proportional improvement in PAB levels. Conclusions: Although the complex interplay between PAB and inflammation cannot be completely solved, we suggest that in subjects receiving an adequate nutritional support a significant improvement of PAB can be considered, at least partially, a real marker of nutrional improvement.