Whether iron metabolism affects metabolic syndrome (METS) is debated. We assessed the association between several markers of iron metabolism and incidence of METS.Data from 3271 participants (1870 women, 51.3 ± 10.4 years), free of METS at baseline and followed for 5.5 years. The association of serum iron, ferritin and transferrin with incident METS was assessed separately by gender. Incidence of METS was 22.6% in men and 16.5% in women (p < 0.001). After multivariate adjustment, a positive association was found between transferrin and incident METS in men: odds ratio (OR) and 95% confidence interval for the fourth relative to the first quartile 1.55 (1.04-2.31), p for trend = 0.03, while no association was found for iron OR = 0.81 (0.53-1.24), p for trend = 0.33 and ferritin OR = 1.30 (0.88-1.92), p for trend = 0.018. In women, a negative association was found between iron and incident METS: OR for the fourth relative to the first quartile 0.51 (0.33-0.80), p for trend<0.03; the association between transferrin and incident METS was borderline significant: OR = 1.45 (0.97-2.17), p for trend = 0.07 and no association was found for ferritin: OR = 1.11 (0.76-1.63), p for trend = 0.58.Transferrin, not ferritin, is independently associated with an increased risk of incident METS; the protective effect of iron in women should be further explored.