Background: Pegylated-interferon plus Ribavirin is the most effective combined treatment for patients with chronic hepatitis C infection. However, response rate in naive patients, irrespectively of genotype, is only about 50%. Aim of this study is to characterize gene expression profiles in naive patients with chronic hepatitis C infection, to investigate the molecular basis for treatment failure. Methods: we collected percutaneous needle biopsy specimens from 20 naive patients who underwent treatment with pegylated interferon a2b plus Ribavirin, and from 3 healthy subjects. Differential gene expression was assessed between three groups with different response to treatment: (A) responders (9), (B) non-responders (7), and (C) relapsers (4). The control group was used to assess gene changes associated with chronic hepatitis C. Results: we identified a predictor set of 90 genes (p < 0.05, of which 17 with p 0.01) whose expression patterns differentiated groups with diferent response after a course of standard regimen (group A vs group B). Moreover gene expression protiling contirmed a trend showing an intermediate profile of group C among group A and B. The comparison of average signal of differentially expressed genes exhibited at least a twofold difference in expression and a false discovery rate of less than I%. Between genes with high expression protile (p 0.01) in responder patients, we identified at least one lSGs gene (Interferon-Stimulated genes) that is completely absent in non-responder patients. Moreover we observed an enhanced expression in group A of 4 additional genes involved directly in antiviral response (p i 0.01 ): inflammatory response, ubiquitination, antigen presentation, T helper 2 signaling pathway and apoptosis. One of the genes with high expression in group B and low in group A is directly involved in multidrug resistance. An other one of those expressed (p <0.01) in non-responders is essential in regulating HCV entry and its replication in the cells. In conclusion we identified some significant signature changes associated with different treatment outcomes that - if confirmed increasing sample size and by Real-Time PCR analysis - could be used in the clinical practice to predict response to a standard come regimen.