To determine whether available lipid-modifying medication can increase high-density lipoprotein (HDL) cholesterol in well-defined genetic or familial HDL-deficiency states, we studied 19 men with HDL deficiency (HDL cholesterol <5th percentile for age and gender) 55 +/- 10 years of age. Concomitant risk factors included diabetes (n = 3) and hypertension (n = 7) and 8 patients had coronary artery disease. Molecular analysis revealed that 4 patients had a mutation in the ABCA1 gene. Patients were assigned to sequentially receive atorvastatin 20 mg/day, fenofibrate 200 mg/day, and extended-release niacin 2 g/day for 8 weeks, with a 4-week washout period between each treatment. Patients in whom a statin was required, according to current treatment guidelines, were kept on atorvastatin throughout the study. Baseline HDL cholesterol level was 0.63 +/- 0.12 mmol/L (24 +/- 5 mg/dl), triglycerides 2.01 +/- 0.98 mmol/L (180 +/- 86 mg/dl), and low-density lipoprotein (LDL) cholesterol 2.29 +/- 0.95 mmol/L (94 +/- 39 mg/dl). Mean percent changes in HDL cholesterol on atorvastatin, fenofibrate, and niacin were -6% (p = NS), +6% (p = NS), and +22% (p <0.05), respectively. Furthermore, niacin significantly increased the large alpha-1 apolipoprotein A-I-containing HDL subspecies (12 to 17 nm). In conclusion, niacin was the only effective drug to increase HDL cholesterol. The absolute increase in HDL cholesterol, approximately 0.10 mmol/L (3.9 mg/dl), is of uncertain clinical significance. Biomarkers of HDL-mediated cellular cholesterol efflux were not changed by niacin therapy. Atorvastatin or fenofibrate had little effect on HDL cholesterol; atorvastatin decreased the total cholesterol/HDL cholesterol ratio by 26%. Fenofibrate did not change HDL cholesterol levels and caused an increase in LDL cholesterol. Aggressive LDL cholesterol lowering may be the strategy of choice in such patients.