When the 'lipid nephrotoxicity hypothesis' was proposed in 1982, it brought together several disparate experimental findings in hyperlipidemia and renal disease to suggest that concomitant hyperlipidemia and proteinuria would cause self-perpetuating renal disease once the initial glomerular insult was no longer present. This process would be analogous to atherosclerosis. Since 1982, increasing evidence has supported the hypothesis that lipid abnormalities contribute to both atherosclerosis and glomerulosclerosis. In this Review, we discuss research developments that are relevant to the lipid nephrotoxicity hypothesis. We describe how inflammatory stress accompanying chronic kidney disease modifies lipid homeostasis by increasing cholesterol uptake mediated by lipoprotein receptors, inhibiting cholesterol efflux mediated by the ATP-binding cassette transporter 1 and impairing cholesterol synthesis in peripheral cells. As a result of these events, cholesterol relocates to and accumulates in renal, vascular, hepatic and possibly other tissues. The combination of increased cellular cholesterol influx and reduced efflux causes injury in some tissues and lowers the plasma cholesterol level. In addition, inflammatory stress causes a degree of statin resistance via unknown mechanisms. These phenomena alter traditional understanding of the pathogenesis of lipid-mediated renal and vascular injury and could influence the clinical evaluation of renal and cardiovascular risk and the role of lipid-lowering treatment in affected patients.