Blood lipids are important biomarkers of risk of coronary heart disease (CHD), the leading cause of death in the world. Myriad data support a causal role of low-density lipoprotein cholesterol (LDL-C) in increasing risk of CHD. Long-standing epidemiology suggests that high-density lipoprotein cholesterol (HDL-C) may protect from disease while high triglycerides (TGs) increase CHD risk. However, the causality of HDL-C and TG to CHD remains controversial. New genetic methodologies have allowed a better look into causal pathways underlying relationships between these traits and disease. Using a combination of approaches for interrogating rare genetic variation in humans, we investigated how HDL and TG may relate to CHD. First, through sequencing and exome-wide genotyping of subjects with extremely high HDL-C, we identified the first homozygote for a loss-of-function (LOF) variant in SCARB1, which encodes scavenger receptor class BI (SR-BI), a hepatic receptor for HDL-C. Despite markedly elevated HDL-C, carriers of this variant had an increased risk of CHD. These findings suggest that HDL functionality in driving cholesterol removal through SR-BI (the reverse cholesterol transport hypothesis) is protective from CHD in humans. Next, we functionally examined one of the first novel loci from genome-wide association studies (GWAS) for HDL-C, GALNT2. Through discovery of humans with genetic GALNT2 LOF and additional studies in rodents and nonhuman primates, we showed that GALNT2 LOF lowers HDL-C across mammals. We also identify one physiological mechanism linking GALNT2 to HDL-C through its enzymatic function. Thirdly, we studied the mechanism of protection of the APOC3 A43T variant recently reported to lower TGs and CHD risk from exome sequencing. Studies in human carriers and animal models suggest that A43T accelerates renal clearance of circulating ApoC-III, thus hindering its function in delaying TG-rich lipoprotein turnover. These data establish ApoC-III clearance mechanisms as potential therapeutic targets for TG lowering. Finally, we adapted a targeted sequencing approach to increase discovery of causal rare coding and noncoding variants at candidate loci influencing HDL-C and TGs. Collectively, this work provides a sampling of approaches for leveraging the spectrum of genomic methods to identify clinically relevant variants impacting HDL, TG and CHD risk.