Endometriosis is a common chronic inflammatory gynaecological condition associated with a spectrum of clinical presentations including different types of pelvic pain and infertility. It is defined as the presence of endometrial-like tissue outside the uterus. Diagnosis requires invasive surgery, resulting in a typical diagnostic delay of 7-10 years, whilst treatment options are limited to repeated surgeries or hormonal medication. Although the disease has a moderate (50%) heritability, the genetic and epigenetic causes of the disease are largely unknown; where genetic variants have been identified, the biological pathways that are perturbed remains unknown. Despite the evidence that endometriosis is a clinically and aetiologically heterogeneous condition, most studies to date are studying endometriosis as binary condition, without regard for the wide spectrum of sub-phenotypes that patients are experiencing. Understanding biological differences between sub-phenotypes of endometriosis, both surgical and symptomatic, is important for the development of better targeted treatment methods as well as non-invasive methods of diagnosis. This thesis aims to examine for the first time differences in genomic, transcriptomic, and epigenomic characteristics of different sub-phenotypes of endometriosis, to aid understanding of their pathogenesis. To improve understanding of 46 established genetic variants associated with endometriosis in genome-wide association studies (GWAS) at a sub-phenotypic level, a genetic association study of detailed phenotypic data from patients recruited in the Oxford ENDOX study was conducted (Chapter 3). Out of 46 GWAS variants for endometriosis, 16 showed nominal association with at least one sub-phenotype: rs11674184/2p25.1/GREB1 and rs1903068/4q12/KDR were associated with ovarian endometriosis (endometrioma), rs71575922/6q25.1/SYNE1 with ASRM stage I/II, and rs12441483/15q15.1/BMF, rs71575922/6q25.1/SYNE1 and rs1903068/4q12/KDR with multiple pain sub-phenotypes. Next, prior to conducting transcriptome and DNA methylome analyses of eutopic endometrium - the tissue of origin for endometriotic lesions - differences in RNA-seq gene expression profiles of eutopic endometrium of ENDOX recruits between major menstrual cycle phases was investigated (Chapter 4) to allow for appropriate adjustment of menstrual cycle phase in subsequent analyses. The analyses confirmed the literature reports of a significant cycle phase effect on gene expression patterns, and highlighted the complement system as one of the most robustly enriched and up-regulated pathways in the secretory phase of the endometrium compared to proliferative phase, which may aid future building of model for cycle phase prediction through transcriptomic profiling. Transcriptomic differences in eutopic endometrium comparing endometriosis cases, by surgical sub-phenotypes, and endometriosis-free controls, were analysed adjusting for cycle phase (Chapter 5). The differential expression analyses in proliferative and secretory phase endometrium highlighted the genome region of 21q11.2 for its potential involvement in inflammatory and oestrogen pathways for ASRM surgical stage I/II and III/IV cases. Five functionally similar clusters of differential expressed genes were proposed: innate immunity (BPIFB1, MUC5AC, MUC5B, EDN2, THBS1), inflammation (SAA1, SAA2), neuro-proliferation (FAT2, FAT3), haemoglobin (HBG2, ALAS2), and lipid metabolism (PLIN1, PLIN4, GPD1). Lastly, differential DNA methylation of eutopic endometrium was explored, comparing endometriosis cases with endometriosis-free controls, in the context of GWAS loci and expression QTL SNPs nearby (in cis) (Chapter 6). For the first time, these analyses considered surgical sub-phenotypes (ASRM disease stage) of endometriosis. The analysis revealed loci at chromosome 4q35.1 (FAM149A & TLR3) that may be responsible for normal regulation of menstrual cycle. Nearby genes of nominally significant EWAS loci (p<1x10-5) for endometriosis suggested five biological mechanisms in common between ASRM surgical stage I/II and III/IV, involving cell cycle (CCND1, RAD52, PAGR1, STEAP3), immunity (PIANP, SPP1, TLR3, SHE), neural regulation (PIANP, C1QL2, GPR158), iron (LYRM4, BOLA2, STEAP3), and oestrogen pathways (PAGR1, ABCG2, SPP1, TLR3, TDRD10). Further validation of the results is needed in an independent dataset before drawing further conclusions. In conclusion, the thesis outlined potential differences between endometriosis sub-phenotypes in genetic, transcriptomic, and epigenetic signatures, although the relatively modest sample sizes mean all results require replication. Nevertheless, they encourage future research to continue exploring sub-phenotypes of endometriosis comprehensively, at much larger scale, using standardised phenotype definitions.