BACKGROUND--Patients with coronary artery disease (CAD) commonly have low HDL cholesterol (HDL-C) and mildly elevated LDL cholesterol (LDL-C), leading to uncertainty as to whether the appropriate goal of therapy should be lowering LDL-C or raising HDL-C. METHODS AND RESULTS--Patients in the Lipoprotein and Coronary Atherosclerosis Study (LCAS) had mildly to moderately elevated LDL-C; many also had low HDL-C, providing an opportunity to compare angiographic progression and the benefits of the HMG-CoA reductase inhibitor fluvastatin in patients with low versus patients with higher HDL-C. Of the 339 patients with biochemical and angiographic data, 68 had baseline HDL-C <0.91 mmol/L (35 mg/dL), mean 0.82+/-0.06 mmol/L (31. 7+/-2.2 mg/dL), versus 1.23+/-0.29 mmol/L (47.4+/-11.2 mg/dL) in patients with baseline HDL-C >/=0.91 mmol/L. Among patients on placebo, those with low HDL-C had significantly more angiographic progression than those with higher HDL-C. Fluvastatin significantly reduced progression among low-HDL-C patients: 0.065+/-0.036 mm versus 0.274+/-0.045 mm in placebo patients (P=0.0004); respective minimum lumen diameter decreases among higher-HDL-C patients were 0. 036+/-0.021 mm and 0.083+/-0.019 mm (P=0.09). The treatment effect of fluvastatin on minimum lumen diameter change was significantly greater among low-HDL-C patients than among higher-HDL-C patients (P=0.01); among low-HDL-C patients, fluvastatin patients had improved event-free survival compared with placebo patients. CONCLUSIONS--Although the predominant lipid-modifying effect of fluvastatin is to decrease LDL-C, patients with low HDL-C received the greatest angiographic and clinical benefit.