The prevalence of iron deficiency was determined in Göteborg, Sweden, in a sample of 15- to 16-y-old girls (n = 220) and boys (n = 207) using serum ferritin (SF). In a recent study in women on the relationship between SF and stainable bone marrow iron, it was established that at a cutoff value for SF of < 16 micrograms/L in 75% of women with no iron stores SF concentration was below this value (sensitivity 75%), whereas in 98% of iron-replete women it was above this cutoff value (specificity 98%). The present study showed that in 40% of the girls and 15% of the boys SF was below this cutoff value, indicating iron deficiency. Low SF concentration was associated with significant decreases in transferrin saturation, Hb concentration, mean corpuscular Hb, and mean corpuscular volume. The results from this cross-sectional study showed that, with decreasing SF, the decrease of values for these parameters occurred already before SF had reached the level 16 micrograms/L, suggesting that SF can be validly used as a single criterion of iron deficiency. Using the cutoff value SF < 16 micrograms/L, the figures for the prevalence of iron deficiency in adolescents in different countries were compared and found to be rather similar in Australia, Canada, the United States, and Sweden. High iron requirements combined with the present low-energy life-style leading to an insufficient supply of dietary iron may be a reasonable main explanation for the paradoxical, high prevalence of iron deficiency in adolescents in affluent societies.