Background: The development of non-invasive procedure to determine HER2 status may represent a powerful method for monitoring disease progression and response to the treatment.Methods: Serum samples and RNA from peripheral blood were evaluated in 85 breast cancer patients (49 HER2 positive and 36 HER2 negative) and 22 healthy controls. HER2 mRNA levels were measured by real-time quantitative PCR (QPCR) and serum HER2 protein by immunoenzimatic assay (EIA). ROC curve analyses were used to determine the optimal cut off values.Results: A statistically significant difference was detected for both QPCR and EIA in HER2 positive patients as compared with both healthy controls and HER2 negative tumours. QPCR showed a 91% (CI95%: 84%–98%) specificity and a 78% (CI95%: 68%–88%) sensitivity for an optimal cut off value of 4.74. The optimal cut off value for EIA was 22 ng/ml yielding a 95% (CI95%: 90%–100%) specificity and a 59% (CI95%: 48%–70%) sensitivity. The QPCR assay was slightly less specific than EIA in discriminating HER2 positive breast cancers from HER2 negative tumours (78% CI95%: 69%–87% versus 86% CI95%: 79%–93%), but it was more sensitive (76% CI95%: 67%–85% versus 55% CI95%: 44%–66%).Conclusions: Our results indicate that QPCR performs better than EIA in the determination of HER2 status of breast cancer patients and could be useful in monitoring the disease during follow up.