This study aimed to evaluate the relationship between low-density lipoprotein cholesterol (LDL-C) levels and cardiovascular disease (CVD) in high-risk patients with hypercholesterolemia without a history of CVD. Patients who were receiving or started treatment with pravastatin, were followed-up for 2 years. Patients were divided into quartiles according to on-treatment LDL-C. The maximum contrast method based on the Cox proportional hazards model was used to evaluate the relationship between achieved LDL-C and the incidence of CVD. Incidence of CVD was also compared according to whether a number of risk factor targets were achieved. A total 6,229 patients were enrolled, with 4,916 having reported LDL-C values. During the 2 years, 69 cases of CVD (6.7/1000 patient years), including 36 coronary artery disease (CAD) (3.5/1000 patient years) and 28 strokes (2.7/1000 patient years), occurred. The comparison of on-treatment LDL-C level quartiles suggested that the incidence of all CVD decreased linearly as the LDL-C levels decreased. Incidence of CAD showed a curvilinear relationship to LDL-C levels, suggesting some attenuation of risk below LDL-C of 119 mg/dL. The incidence of all CVD and CAD tended to be decreased as the number of achieved risk factor targets increased. In conclusion, through our observational study, it was shown that a linear relationship between the incidence of CVD and LDL-C was observed in high-risk hypercholesterolemic patients. The low incidence of CVD in the present study may be associated with multifactorial management of conventional risk factors including high LDL-C levels. However, prospective, randomized studies are needed to confirm these findings.