Objective: To investigate the prevalence of iron deficiency among schoolchildren of different socio-economic status (SES), living in the three largest cities of Turkey. Design: Cross-sectional study. Settings: Primary schools of Istanbul, Ankara and Izmir. Subjects: Schoolchildren aged 12 and 13 y (males: 504; females: 510) from nine primary schools. Inclusion of subjects in the study was on a voluntary basis. Methods: Data were obtained on children SES, anthropometry, haematological and biochemical indices of iron status and consumption of food items related to dietary iron bioavailability. One-way analysis of variance was mainly applied, for the evaluation of the tested hypotheses, using Bonferroni correction in order to take into account the inflation of Type I error. Results: Iron deficiency (serum ferritin o15 mg/l) prevalence was 17.5% among boys and 20.8% among girls. Furthermore, iron deficiency was significantly more prevalent among boys of lower SES, who were also found to have significantly lower levels of serum iron, serum ferritin, transferrin saturation, mean corpuscular haemoglobin and mean corpuscular haemoglobin concentration compared to those of higher SES. In terms of dietary factors affecting iron bioavailability, low SES boys exhibited significantly higher frequency of tea consumption and lower frequency of citrus fruit, red meat and fish consumption, compared to their higher SES counterparts. Conclusion: The prevalence of iron deficiency was relatively high, particularly among lower SES schoolboys. Higher tea and lower citrus fruits, red meat and fish consumption by boys of lower SES may provide an indication about the possible role of certain dietary patterns in the different manifestation of this medical condition among the socio-economic groups. However, further research is needed.