Of the ~3.3 billion bases of the human genome, only about 2% code for proteins and until very recently, the remaining 98% have been considered to be 'junk'. However, large transcriptomic studies have shown that around 90% of the genome is actively transcr ibed of which a significant fraction may contribute to a previously underestimated layer of regulatory long non-coding RNAs (lncRNAs). In several studies it has been shown that the expression of small ncRNAs, like microRNAs, is associated with diseases in cluding cancer. However, the large group of long lncRNAs has drawn less attention. LncRNA expression of selected 26 breast carcinoma samples, representing the five clinically relevant tumor expression subclasses, and 5 samples from normal breast tissue we re analyzed using the nONCOchip®. The nONCOchip®, developed by the RNomics group at the Fraunhofer Institute for Cell Therapy and Immunology, covers both, experimentally identified cancer related lncRNAs as well as known or predicted non-coding RNAs from public databases. The array includes in total 243,000 probes, with over 60,000 newly identified transcripts. First results reveal that lncRNAs are variably expressed in breast tissue, expression signatures are different between breast tumors and normal br east tissues, expression patterns in breast cancer show a high degree of heterogeneity, and certain lncRNAs are significantly differentially expressed, e.g. between TP53 wild type and mutated samples. The overall aims of this study are to identify lncRNAs involved in gene regulation in breast cancer, to explore lncRNAs gene aberrations/expression profiles important for breast cancer development and progression, and to identify lncRNAs biomarkers predictive for breast cancer. As the growing list of lncRNA genes influencing carcinogenesis is striking, the identification of clinically relevant lncRNAs may open up for the development of novel biomarkers and therapeutic tools that attack diseased cells.