Objectives: We investigated the prevalence of scoliosis among primary school students in Sivas, Turkey. Methods: To determine the prevalence of scoliosis among primary school students in the age bracket of 10 to 15 years, 11 primary schools were selected with systematic sampling. A total of 3,175 students (1,538 girls, 1,637 boys) of grades 6 to 8 were selected from 16,103 students using a stratified sampling method. Examination for scoliosis was made by the forward bend test and palpation of the spine. Results: Fifteen children (0.47%) were found to have scoliosis, including 10 girls (66.7%) and five boys (33.3%). The prevalence of scoliosis was significantly higher in girls (0.65% vs. 0.31%; p<0.05). The mean lateral curvature of the spine was 6.9° (range 5° to 20°), being 7.9±4.6° in girls, and 5.4±0.9° in boys. The mean age was 13.5±1.2 years (range 12-15 years). The severity of the curvature showed no significant relationship with gender and age groups (p>0.05). The levels of involvement were lumbar vertebrae in 73.3% (n=11), thoracic vertebrae in 13.3% (n=2), and thoracolumbar vertebrae in 13.3% (n=2). Spinal curvature was to the right side in 12 cases (80%), and to the left in three cases (20%). Girls and boys did not differ significantly with respect to the severity and direction of the curvature (p>0.05). During a two-year follow-up of children with scoliosis, no progression of the curvature was observed, including one child who wore a Milwaukee brace. Conclusion: School screenings may be performed as part of prevalence studies; however, routine school screening for scoliosis is debatable.