We investigated changes in the prevalence of myopia and mean changes in refractive errors in Japanese students from 3 to 17 years old from 1984 to 1996. Mass ophthalmologic surveys were performed annually during the course of the study. The age-specific frequency distribution of refractive errors remained similar for 6-year-old students (defined in this study as students in the first grade of primary school) during the 13-year period, but the distribution became gradually skewed toward myopia for 12-year-old students (defined in this study as students in the first grade of junior high school). Comparisons between 1984 and 1996 examinations showed a considerable increase in the incidence of myopia among those 7 years of age or older, and changes in mean refractive errors also demonstrated a greater shift toward myopia, especially in students older than 10 years, for whom the changes were statistically significant. In this 13-year period, the prevalence of myopia increased from 49.3% to 65.6% in 17-year-old students. In addition to the annual mass ophthalmologic examinations, we also performed a longitudinal 6-year study of 346 students who entered junior high school in 1989, 1990, or 1991. Among these students, the prevalence of myopia increased from 43.5% at 12 years of age to 66.0% at 17 years of age. These 346 students were divided into the following eight groups according to their refractive error (spherical power [D]) at 12 years of age: +1 D, 0 D, -1 D, -2 D, -3 D, -4 D, -5 D, and -6 D. Mean progressions of myopia in these students were as follows: for the +1 D group, -0.14 D/year; for the 0 D group, -0.25 D/year; for the -1 D group, -0.37 D/year; for the -2 D group, -0.40 D/year; for the -3 D group, -0.29 D/year; for the -4 D group, -0.25 D/year; for the -5 D group, -0.14 D/year; and for the -6 D group, -0.22 D/year. Boys and girls demonstrated a statistically significant difference in mean changes in refractive errors at the 6-year follow-up examination: the mean change in refractive error was -1.41 +/- 1.25 D for boys as compared with -1.03 +/- 1.07 D for girls (unpaired Student's t-test, P < 0.0001). Our results demonstrated an early age at onset for myopia and a recent increase in the proportion of myopic students. Further studies are needed to shed light on the extent to which myopia is caused by environmental factors, because it is through these factors that the prevalence rate may be affected.