In 1996, the first 2 studies using 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitor ("statin") therapy in hypertriglyceridemic subjects were published. In subjects with isolated triglyceride elevations who were treated with atorvastatin 5, 20, and 80 mg/day, large and dose-related reductions were noted. In subjects with combined hyperlipidemia treated with 10 mg simvastatin, triglyceride reduction similar to that reported for the 5 mg atorvastatin dose was seen. In response to these findings, we conducted comparative assessments to determine whether all statins are effective in lowering triglyceride levels and whether their effect on triglycerides is related to factors such as drug, dose, and baseline triglyceride levels. To standardize these assessments, we devised a ratio that related changes in triglyceride levels to the known predictable response of low-density lipoprotein (LDL) cholesterol to statins. This triglyceride/LDL cholesterol ratio was obtained by dividing the percent change from baseline in the triglyceride level by the percent change from baseline in the LDL cholesterol level. The triglyceride/LDL cholesterol ratio was initially applied to several published studies, and found to be approximately 1.0 and 0.5 in hypertriglyceridemic and nonhypertriglyceridemic populations, respectively. We then assessed the effect of various statins on triglycerides using a pooled laboratory database of 2,689 subjects who had participated in 7 separate studies with similar designs. All of the studies had a placebo run-in followed by a randomized, double-blind, active treatment phase of at least 4 weeks with a statin. Entry into these studies required a triglyceride level of <400 mg/dL. In subjects with baseline triglyceride >250 mg/dL, significant and dose-dependent reductions in triglyceride of 22-45% were seen with all statins. When baseline triglyceride was <150 mg/dL, no significant or dose-dependent effect on triglyceride was seen. The triglyceride/LDL cholesterol ratio was evaluated using a linear model that included baseline triglyceride level, drug, and dose. Only the baseline triglyceride level was significantly (p <0.001) related to this ratio. Moreover, the triglyceride/LDL cholesterol ratio was fairly constant across all statins and doses for patients with baseline triglyceride levels of <150 mg/dL, 150-250 mg/dL, and >250 mg/dL, at 0.0+/-0.3, 0.5+/-0.2, and 1.2+/-0.3, respectively. We conclude that all statins are effective in decreasing triglyceride levels, but only in hypertriglyceridemic patients. Due to the relatively constant triglyceride/LDL cholesterol ratio, our analysis indicates that the more effective the statin is in decreasing LDL cholesterol, the more effective it will also be in decreasing triglyceride levels in patients with hypertriglyceridemia.