Lipids play critical roles in several major chronic diseases of our times, including those that involve inflammatory sequelae such as metabolic syndrome including obesity, insulin sensitivity, and cardiovascular diseases. However, defining the substrate specificity of enzymes of lipid metabolism is a challenging task. For example, phospholipase A2 (PLA2) enzymes constitute a superfamily of degradative, biosynthetic, and signaling enzymes that all act stereospecifically to hydrolyze and release the fatty acids of membrane phospholipids. This review focuses on how membranes interact allosterically with enzymes to regulate cell signaling and metabolic pathways leading to inflammation and other diseases. Our group has developed "substrate lipidomics" to quantify the substrate phospholipid specificity of each PLA2 and coupled this with molecular dynamics simulations to reveal that enzyme specificity is linked to specific hydrophobic binding subsites for membrane phospholipid substrates. We have also defined unexpected headgroup and acyl chain specificity for each of the major human PLA2 enzymes, which explains the observed specificity at a structural level. Finally, we discovered that a unique hydrophobic binding site-and not each enzyme's catalytic residues or polar headgroup binding site-predominantly determines enzyme specificity. We also discuss how PLA2s release specific fatty acids after allosteric enzyme association with membranes and extraction of the phospholipid substrate, which can be blocked by stereospecific inhibitors. After decades of work, we can now correlate PLA2 specificity and inhibition potency with molecular structure and physiological function.