The one-step nucleic acid amplification (OSNA) method is an increasingly used procedure for intraoperative analysis of sentinel lymph node (SLN) status in breast cancer patients. It measures cytokeratin19 (CK19) mRNA copy numbers in homogenized samples of SLN; CK19 has been chosen for identifying node metastasis because most breast cancers express this molecule. However, to avoid false-negative OSNA results, testing the preoperative needle core biopsy (NCB) of breast carcinomas for CK19 by immunohistochemistry (IHC) has been recommended. This procedure relies on the assumption that protein expression is strictly related to mRNA expression. We developed this study to evaluate if IHC gives similar result to the molecular assay. In a series of breast cancer patients with axillary metastasis, corresponding surgically resected tumor and metastatic lymph node specimens have been tested for CK19 protein by IHC and for CK19 mRNA by real-time PCR; furthermore, CK19 immunostaining has been performed in NCBs when available. Statistical analysis revealed that (1) the immunohistochemical evaluation of CK19 in NCB is a reliable test, reflecting protein expression in the whole tumor and in the metastatic lymph node; (2) there is no correlation between CK19 protein expression and CK19 RNA level neither within the primary breast cancer nor within the metastatic node; moreover, no correlation as well has been found between protein expression in NCB and mRNA level in metastatic lymph nodes. Thus, our results suggest that there is no evidence-based reason to stain every NCB for CK19 before performing OSNA in patients with breast cancer.