Molecular profiling of breast cancer has gained popularity due to the possibility of studying the biological spectrum of the disease and to evaluate prognostic and predictive characteristics that can lead to more accurate therapeutic decisions. Although studies have been carried out on tissue using tissue microarray technology (TMA), to our knowledge, this technology has not been applied to samples obtained by fine-needle aspiration biopsy (FNAB). This study was designed to correlate FNAB samples with corresponding surgical specimens for the assessment of basal phenotype in patients with triple negative breast cancer (TNBC). A total of 198 cases of TNBC with matching FNAB and surgical specimens were identified. Forty-six cases with sufficient tissue in both FNAB cell blocks and surgical specimens were selected. Tissue microarray blocks were prepared and stained with six biomarkers (CK5/6, p63, SMA, EGFR, C-Kit, and p53). For statistical analysis, we used the observed (Gross) percentage of agreement and also calculated the Cohen's Kappa for each biomarker. We found a high agreement between the two groups as shown by the values of Observed (Gross) percentage of agreement (mostly 90% or higher except for the C-Kit which was 78%). In addition, Cohen's Kappa point estimate showed substantial agreement (0.61–0.80) for CK 5/6, p63, EGFR, p53, moderate agreement (0.41–0.60) for C-Kit, and, fair agreement (0.21–0.40) for SMA. Our study shows that the advantages of FNAB are not restricted to cost-effectiveness, but also attested that cytology samples are suitable for the evaluation of biomarkers that have important implications on patient's therapy and prognosis. Diagn. Cytopathol. 2011;39:1–7. © 2010 Wiley-Liss, Inc.