PURPOSE: To determine the rate of underestimation of an image-guided core biopsy of nonpalpable breast lesions, with validation by histologic examination after surgical excision. METHODS: We retrospectively reviewed 352 biopsies from patients who were submitted to surgery from February 2000 to December 2005, and whose histopathologic findings were recorded in the database system. Results were compared to surgical findings and underestimation rate was determined by dividing the number of lesions that proved to be carcinomas at surgical excision by the total number of lesions evaluated with excisional biopsy. Clinical, imaging, core biopsy and pathologic features were analyzed to identify factors that affect the rate of underestimation. The degree of agreement between the results was obtained by the percentage of agreement and Cohen's kappa coefficient. The association of variables with the underestimation of the diagnosis was determined by the chi-square, Fisher exact, ANOVA and Mann-Whitney U tests. The risk of underestimation was measured by the relative risk (RR) together with the respective 95% confidence intervals (95%CI). RESULTS: Inconclusive core biopsy findings occurred in 15.6% of cases. The histopathological result was benign in 26.4%, a high-risk lesion in 12.8% and malignant in 45.2%. There was agreement between core biopsy and surgery in 82.1% of cases (kappa=0.75). The false-negative rate was 5.4% and the lesion was completely removed in 3.4% of cases. The underestimation rate was 9.1% and was associated with BI-RADS® category 5 (p=0,01), microcalcifications (p CONCLUSIONS: The core breast biopsy under image guidance is a reliable procedure but the recommendation of surgical excision of high-risk lesions detected in the core biopsy remains since it was not possible to assess clinical, imaging, core biopsy and pathologic features that could predict underestimation and avoid excision. Representative samples are much more important than number of fragments.