I investigated genetic and epigenetic variants underlying obesity and related metabolic disturbances with the aim of identifying new disease mechanisms. I first examined the contribution of genetic variation to the increased risk of central obesity (waist-hip-ratio, WHR) in South Asians compared to Europeans by: i) genome-wide association in 10,318 South Asians; ii) exome-wide association in 2,637 South Asians; iii) comparison of risk allele frequencies and effect sizes of 48 known WHR SNPs in South Asians and Europeans. I found no associations between population-specific or cosmopolitan SNPs and WHR, and no evidence of larger genetic effects at known WHR loci in South Asians. I next investigated the cell lineages and causal pathways underlying DNA methylation markers associated with BMI in blood (187 loci, P < 1x10-7) by: i. replication testing the loci in isolated white blood cells and adipocytes from obese cases and controls; ii. analysing the relationships between methylation, adiposity, and metabolic disturbances using genetic association. I showed that DNA methylation in blood reflects methylation in isolated white cells (enrichment P < 1.2x10-9) and adipocytes (enrichment P < 0.05), and is predominantly a consequence rather than a cause of adiposity. I finally evaluated genome-wide DNA methylation in isolated subcutaneous and visceral adipocytes from 24 morbidly obese cases and 24 controls (Illumina-450K) to identify putative obesogenic and diabetogenic processes. I discovered widespread differences in DNA methylation between: i. cases and controls (140 loci, P < 1x10-7); ii. subcutaneous and visceral adipocytes (671 loci, P < 1x10-7). Loci were enriched in multiple cellular/molecular pathways (case-control N=47, subcutaneous-visceral N=84, P < 0.001), and identified known obesity and metabolic genes (e.g. IRX3/IRX5, TBX15, PRMD16). These discoveries will be replicated and functionally characterised in future work. Overall, my results suggest that DNA methylation in adipocytes may be important in obesity and metabolic disease pathogenesis, and that methylation in blood may serve as a biomarker of tissue-specific pathobiological processes.