Little is known about the efficacy of left ventricular hypertrophy diagnosed by electrocardiography for predicting cardiovascular disease in a general Japanese population. In a large cohort of participants selected randomly from the overall Japanese population, we attempted to evaluate the usefulness of a high amplitude R-wave (left high R-wave) on the electrocardiogram for predicting cardiovascular death. A total of 6,688 Japanese (mean age, 50.7 years old; 57% women) free of previous cardiovascular disease and use of antihypertensive agents at baseline were followed for 10 years, from 1990 to 2000. Left high R-wave on the electrocardiogram (the Minnesota Code, 3-1 or 3-3) was found in 9.4% of the 6,688 participants, in 14.6% of the 2,413 hypertensives and in 6.4% [corrected] of the 4,275 normotensives. During the follow-up period, 133 [corrected] participants died due to cardiovascular disease. After adjustment for systolic blood pressure and other risk factors, left high R-wave conferred an increased risk of cardiovascular death; the hazard ratio among all the participants was 1.88 (95% confidence interval, 1.22-2.89; p < 0.01), that among hypertensives was 1.97 (1.20-3.24; p = 0.01), and that among normotensives was 1.66 (0.69-3.98; p = 0.26). The population attributable risk percent of left high R-wave for cardiovascular death was 7.6% among all participants, 12.4% among hypertensives and 4.1% among normotensives. Left high R-wave on electrocardiogram, irrespective of the level of systolic blood pressure, was a predictive marker for cardiovascular death among community-dwelling Japanese.