Fludarabine (F-ara-A) is widely used as palliative treatment in chronic lymphocytic leukemia (CLL). Clinical resistance is frequently observed, and adverse effects are common. To date, no practical assay exists to identify patients likely to derive benefit from F-ara-A. We previously reported that high mRNA levels encoding human concentrative nucleoside transporter 3 (hCNT3) protein in CLL correlated with clinical resistance to F-ara-A. This study explores the value of immunohistochemistry (IHC) for hCNT3 as a marker of F-ara-A resistance in CLL. We studied 36 CLL patients who received F-ara-A monotherapy and had suitable pre-F-ara-A tissue available. IHC was performed with validated hCNT3-specific monoclonal antibodies and quantitatively scored by a hematopathologist blinded to clinical outcomes. Relationships between hCNT3 staining in CLL cells and time to progression (TTP), overall response (OR), and overall survival (OS) were assessed. Dichotomization of quantitative hCNT3 staining showed that subjects with high hCNT3 IHC scores had a significantly shorter TTP with F-ara-A treatment compared to those with a low score (hazard ratio, HR, 3.16; P=0.006). Median TTP was 4.7 vs 11.2 months, respectively. On multivariate analysis, hCNT3 score was the only clinical parameter independently associated with TTP (HR, 3.12; P=0.01). OR and OS did not differ significantly between the dichotomized groups. We found a strong relationship between IHC staining of hCNT3 and clinical resistance to F-ara-A therapy in CLL. If confirmed, IHC for hCNT3 may be routinely used to predict those patients unlikely to benefit from F-ara-A, thereby avoiding F-ara-A-related toxicities.