The aim of this study was to develop and validate a food frequency questionnaire (FFQ) on the dietary intake of Turkish school-aged children. Fifty randomly selected students aged 7 – 12 from urban areas of Istanbul were included in this study. An FFQ, containing a list of frequently consumed foods which 138 item was developed. Dietary records (DRs) including three days, and FFQs were collected during autumn and spring. Daily consumption of each food group was assessed and the nutrient compositions of diet were calculated. The Pearson correlation coefficient, weighted kappa, The Bland-Altman scatter plots between averages of the reported (FFQ) and references method (DR) were used as validity coefficient. The Pearson correlation, energy adjusted and attenuation coefficients of nutrient intake were between for energy (0.29; 0.11), protein (0.36 - 0.32; 0.21), carbohydrate (0.25 - 0.09; 0.15), fat (0.25 - 0.05; 0.13). As food groups, for meat group (0.29 – 0.21, 0.09), oil and margarine groups (0.21 – 0.02, 0.08), dairy group (0.52 – 0.58, 0.31), vegetable group (0.14 – 0.11, 0.09), fruit group (0.31 – 0.40, 0.09) and cereal group (0.33 – 0.34, 0.10), respectively. According to the kappa values, there is moderate and fair agreement between two methods. The Bland-Altman scatter plots showed an acceptable level of agreement between the two methods that DRs may be replaced with FFQ. The performance of the FFQ is likely to allow detecting and sufficient assessment of nutritional status. The developed FFQ had validity coefficients similar to those FFQs in previous studies.