The human breast is a complex and dynamic organ that harbours various cell populations. Interestingly, different mammary epithelial cell populations give rise to different breast cancer subtypes and the cell-of-origin strongly influences the tumour molecular characteristics and clinical outcomes. To fully understand the process of tumourigenesis, it is necessary to characterise the complete spectrum of human mammary epithelial cell populations. While the landscape of protein-coding genes has been extensively explored in cancer, lncRNAs (predominantly unannotated) are comparatively unchartered. Using deep bulk RNAseq combined with de novo transcript assembly, we discovered >13,000 novel lncRNAs expressed in normal mammary epithelial cells that are not annotated in existing databases. We then surveyed the transcriptional landscape of these lncRNAs across the different cell subpopulations of the normal breast in very high resolution using single-cell RNAseq data. Notably, clustering cells with Seurat based only on the expression of lncRNAs, we were able to distinguish the main mammary epithelial cell populations. Moreover, we showed that lncRNAs are significantly more cell type-specific than their protein-coding counterpart, suggesting lncRNAs represent a better resource of biomarkers for cell populations. Accordingly, we identified an average of 150 lncRNA markers per cell cluster, which can potentially be used as biomarkers for different breast cell populations. We hypothesised that subsets of lncRNAs would also be specifically expressed in different breast cancer subtypes, reflecting their cell-of-origin. Indeed, comparing the markers for each cell population with molecular signatures of breast cancer subtypes from Metabric and TCGA, we identified lncRNAs implicated in subtype specificity, some of which may be involved in tumourigenesis. In conclusion, by exploiting the unannotated transcriptome, we assessed thousands of noncoding transcripts overlooked by previous studies. Our innovative approach allowed us to identify new markers for normal mammary epithelial cell populations and breast cancer subtypes, as well as potential noncoding oncogenes.