Objective The presence of thyroid antibodies in pregnancy has been associated with preterm birth. In the non-pregnant population, the implementation of the Danish iodine fortification program has increased the prevalence of thyroid antibodies. This study investigated the prevalence of thyroid peroxidase antibodies (TPOAbs) and thyroglobulin antibodies (TgAbs) in pregnant Danish women before, during and after implementation of the iodine fortification program and association with preterm birth. Design Comparative cohort study of 1368 pregnancies from three cohorts gathered before (1996–1998), during (2000–2003) and after (2008–2009) the iodine fortification program. Methods In cohort 1 (n = 297), TPOAbs were measured (DYNOtest (BRAHMS)). In cohorts 2 (n = 148) and 3 (n = 923), both TPOAbs and TgAbs were measured (Kryptor immunofluorescent assay (BRAHMS)). The prevalence and effect of antibody positivity were explored using three cut-offs: TPOAbs and/or TgAbs >100 kU/L, TPOAbs and/or TgAbs >60 kU/L and TPOAbs >30 and/or TgAbs >20 kU/L. National preterm birth data were extracted from the National Birth Registry. Results In the three cohorts, TPOAb levels >60 kU/L were found in 5.4, 8.1 and 12.0% (χ ²(2, n = 1367) = 11.7, P = 0.003) respectively, and TPOAbs and/or TgAbs >60 kU/L in 8.1 and 16.2% in cohorts 2 and 3 respectively (χ ²(2, n = 1070) = 6.5, P = 0.01). TgAb levels (>20 kU/L) had increased plenty-fold from cohort 2 to 3 (χ ²(1, n = 1071) = 136.5, P < 0.001). Preterm birth occurred in 4.1% of all pregnancies with no effect from antibody positivity (TPOAbs and/or TgAbs >60 kU/L, χ²(1, n = 1039) = 0.0, P = 0.98, aOR = 1.1, 95% CI (0.4–2.7)). The national preterm birth-rate showed no increase over the same period. Conclusions Thyroid antibody positivity in Danish pregnant women has more than doubled upon the implementation of the iodine fortification program without an increase in preterm birth-rate.