Objectives: To investigate whether thyroid function profiles can predict the poststroke fatigue (PSF) in patients with AIS. Methods: Stroke patients were consecutively recruited within 3 days of onset in Jinling hospital. Serum levels of thyroid hormones, thyroid antibodies, haematological indexes, and biochemical indexes were measured on admission. Fatigue was scored by the FSS. Association were analyzed with multivariate regression and restricted cubic splines. Results: Of the 704 stroke patients, 292 (41.5%) were diagnosed with fatigue in the acute stage and 224 (35.3%) at 6-month of the index stroke. The serum levels of TSH were inversely associated with the risk of PSF in both acute phase and follow-up evaluations after adjusting for potential confounders (OR 0.30, 95% CI 0.24-0.37 in the acute phase and OR 0.70, 95% CI 0.58-0.84 at follow-up). The subgroup analysis indicated that in acute phase of ischemic stroke, TSH was associated with the severity of PSF in group of euthyroidism (β = -0.70, p<0.001), subclinical hypothyroidism (β = -0.44, p<0.001) and low-T3 syndrome (β = -0.34, p=0.008). Higher TSH was associated with the better FSS scores in patients with low- T3 syndrome at 6-month of the index stroke (β = -0.35, p=0.01). Furthermore, in the group of low-T3 syndrome, FT3 serum level could also indicate a higher risk of PSF (β = -2.54, p<0.001 in the acute phase and β = - 2.67, p<0.001 at follow-up). Conclusions: Thyroid function profiles may predict fatigue after AIS, suggesting that neuroendocrine responses could play a role in PSF.