Aim Breast carcinoma proliferative activity, histological grade and commercial molecular tests are all important in prognostication and treatment. There is a particular need for improved, standardised techniques for subclassification of grade 2 breast cancers into low-risk and high-risk prognostic groups. In this study we investigated whether gene expression profiling of five proliferation genes was feasible using breast cancer tissue in a clinical setting and whether these profiles could enhance pathological assessment. Methods Expression of five proliferation gene mRNAs; Ki-67, STK 15, CCNB1, CCND1 and MYBL2, was quantified in 27 breast carcinomas and compared with Ki-67 proliferation index (PI) and Nottingham mitotic score. Results Expression of Ki-67, STK15 and MYBL2 mRNA showed moderate Spearman's correlation with Ki-67 PI (p<0.01), but CCND1 and CCNB1 showed weak, non-significant correlation. Individual gene expression did not associate with mitotic score but combined mRNA expression correlated with both Ki-67 PI (p=0.018) and mitotic score (p=0.03; 0.007). Conclusions This study confirms mRNA analysis in breast carcinoma formalin-fixed, paraffin-embedded samples is feasible and suggests gene expression profiling, using a small set of five proliferation genes, has potential in aiding histological grading or assessment of proliferative activity of breast cancers. To fully evaluate the clinical applicability of this approach, a larger cohort study with long-term follow-up data is required.