<b><i>Introduction:</i></b> Thyroid function in Rett syndrome (RTT) has rarely been studied with unanimous results. However, this aspect is of great concern regarding the effect thyroid hormones (TH) have on proper mammalian brain development. <b><i>Objective:</i></b> To evaluate the prevalence of abnormalities of thyroid function in a cohort of children with RTT. <b><i>Patients and Methods:</i></b> Forty-five consecutive Caucasian girls (mean age: 8.6 ± 5.3 years, range: 2.0-26.1) meeting the clinical criteria for RTT were recruited. In all of the subjects, we evaluated the serum concentrations of free-T₃ (FT₃), free-T₄ (FT₄), thyroid-stimulating hormone (TSH), thyroperoxidase autoantibodies, thyroglobulin autoantibodies (TgA), and TSH receptor (TSHr) autoantibodies. The results were compared with a group of 146 age-matched healthy Caucasian children and adolescent girls (median age: 9.5 years, range: 1.8-14.6) from the same geographical area. <b><i>Results:</i></b> Mean FT₃ and TSH levels were not significantly different between the RTT patients and controls. Nevertheless, FT₄ levels were significantly higher in RTT patients than in controls (p < 0.005). In particular, 17.7% showed FT₄ levels higher than the upper reference limit (vs. 0.7% of controls, p < 0.0001), whereas 12 patients (26.7%) showed higher FT₃ levels than the upper reference limit, significantly differing in respect to controls (2.0%, p < 0.0001). Finally, 5 patients (11.1%) showed higher levels of TSH, statistically differing from the control subjects (2.0%, p < 0.0001). However, evaluating the patients on the basis of different RTT genotype subgroups, patients with <i>CDKL5</i> deletions showed significantly higher FT₄ values than patients with <i>MeCP2</i> deletions (p < 0.05). On the other hand, patients with other types of <i>MeCP2 </i>mutations also showed FT₄ levels significantly higher than patients with <i>MeCP2 </i>deletions (p < 0.05). In fact, out of 8 patients with FT₄ levels higher than the upper references limit, 3 of them presented with <i>CDKL5</i> deletions (3 patients, 37.5%), 4 (50%) had <i>MeCP2</i> mutations, and 1 (12.5%) belonged to the subgroup of <i>MeCP2</i> deletions. However, when analyzing FT₃ levels of the 12 patients showing higher FT₃ levels than the upper references limit, 6 (50%) belonged to the subgroup with <i>MeCP2 </i>mutations, 4 (33.3%) to the subgroup with <i>MeCP2 </i>deletions, and 2 (16.7%) to the subgroups with <i>CDKL5</i> deletions. Furthermore, no patient with RTT was positive for antithyroglobulin autoantibodies, antithyroid peroxidase, or anti-TSHr, with no statistical differences in respect to the controls. <smlcap>L</smlcap>-thyroxine treatment was not necessary for any patient. <b><i>Conclusions:</i></b> Abnormalities of thyroid function are not rare in RTT. The possible relationship between these disorders and the RTT phenotype should be confirmed and studied. Children with RTT should be screened for potential thyroid dysfunction.