Autotaxin (ATX) is a ubiquitous ectoenzyme that hydrolyzes lysophosphatidylcholine to form the bioactive lipid mediator lysophosphatidic acid (LPA). LPA activates specific G-protein-coupled receptors to elicit downstream effects leading to cellular motility, survival, and invasion. Through these downstream effects, autotaxin is involved in many human disorders including cancer, heart disease, chronic pain, asthma, and other inflammatory diseases. Inhibition of autotaxin activity is therefore a therapeutically attractive goal. This review will summarize insights into the inhibition of ATX gained across multiple structure-activity relationship (SAR) studies. Initial ATX inhibitors included metal chelators such as L-histidine and analogs. Subsequently, the bioactive lipids LPA and sphingosine-1-phosphate were identified as ATX feedback inhibitors, which lead to extensive SAR studies on phospholipid analogs. Subsequently, computational and experimental SAR studies have allowed for the discovery of small, non-lipid ATX inhibitors that, in many cases, avoid G-protein-coupled receptor-mediated off-target effects. Crystal structures of ATX inhibitor complexes have been reported starting from 2011, confirming earlier enzyme kinetic studies that suggested multiple inhibitor binding sites. These structures revealed a unique allosteric hydrophobic pocket within the ATX catalytic domain. Further exploration of this hydrophobic pocket, both computationally and experimentally, has been undertaken to develop additional ATX inhibitors that take advantage of hydrophobic interactions.