Introduction: The aim of the present study was to investigate the reference intervals for thyroid stimulating hormone (TSH) in healthy individuals with normal levels of serum free thyroxine (fT4) and without sonographic pathologies, and determine the effects of age, gender, and residence on the TSH reference intervals. Subjects and methods: This research was a population-based study conducted in 70 regions. The random sampling method was used to select the 1095 subjects of the study among inhabitants aged 18 and above. Patients who had a previous history of thyroid disease and had been taking medication were excluded from the study as this may have affected their fT4 or TSH levels. In addition, subjects who had serum fT4 without a reference range and abnormal ultrasonography findings were also excluded. A total of 408 subjects were used for establishing the reference intervals for TSH. Results: The data for TSH in the study group were not normally distributed according to the Kolmogorov-Smirnov index. The geometric mean was 1.62 mIU/L, the median was 1.40 mIU/L, and the 95% reference intervals were 0.38-4.22 mIU/L. The median TSH level was higher in females compared to males (p < 0.05). In the female subjects 2.5th percentile of TSH was lower and 97.5th percentile was higher than those of males. The reference intervals of TSH were of lower values in subjects over 50 years old (p < 0.001). Discussion: Studies suggest that determination of the TSH reference intervals may differ due to environmental influences or due to age, gender, and race. It is suggested that the lower limit of normal TSH for the adult Turkish population would be 0.38 mIU/L and the upper limit similar to the traditional value of 4.2 mIU/L. If each clinician uses their population-specific reference interval for TSH, thyroid function abnormalities can be accurately estimated.