The efficacy, safety, and tolerability of a moderate dose, 3-drug lipid-lowering regimen were evaluated among 29 male patients with hyperlipidemia and coronary artery disease. In an initial 12-month phase, regular niacin, 500 mg qid, lovastatin, 20 mg bid, and colestipol, 10 g/bid, were given with dose adjustment for lipid targets and side effects. This was followed by 2 random sequence crossover phases (8 months each) alternating regular niacin with a polygel controlled-release formulation of niacin for use in this regimen. Lipid, lipoprotein, apoprotein, and clinical chemistry determinations were obtained at baseline, during the initial phase, at the 2 crossover phases, and at 6 weeks after therapy. A final questionnaire queried specific side effects and overall preferences. Low-/high-density lipoprotein (LDL/HDL) changed from means of 215/46 mg/dl at baseline, to 94/59 mg/dl after run-in, to 85/52 mg/dl after 8 months of controlled-release niacin, and to 98/56 mg/dl after 8 months of regular niacin (regular niacin vs controlled-release niacin, p <0.005/<0.05). The target of LDL < or = 100 mg/dl was achieved at 8 months by 83% of these patients with controlled-release niacin and by 52% with regular niacin (p <0.01). Compliance was 95% with controlled-release niacin versus 85% with regular niacin (p <0.001). The controlled-release niacin and regular niacin regimens did not differ in terms of uric acid, glucose, insulin, or asparate aminotransferase levels. Overall, 21% of patients called the 3 drugs "very easy" and 72% "fairly easy" to take. The controlled-release niacin-containing regimen was preferred by 21 patients and the regular niacin by 4. In conclusion, these regimens achieve striking lipid changes among hyperlipidemic patients. Controlled release is the preferred niacin preparation in terms of LDL reduction, compliance, patient preference, and achieving the National Cholesterol Education Program guideline of LDL < or = 100 mg/dl. The 2 niacin preparations did not differ in evidence of toxicity.