The objective of this study was to evaluate the relationship between subclinical hypothyroidism (SCH) and depression. We also analysed the effect of levothyroxine (L-T4) on depression in SCH patients. We found an insignificant difference for the composite endpoint: standard mean difference (SMD) of 0.23 (95% confidence interval (CI) −0.03, 0.48, P = 0.08, I2 = 73.6%). The odds ratio (OR) for depressive patients was 1.75 (95% CI 0.97, 3.17 P = 0.064, I2 = 64.6%). Furthermore, sub-group analysis according to age found that SCH was related to depression in younger patients (<60 years old), as defined by the diagnosis of depression: OR of 3.8 (95% CI 1.02, 14.18, P = 0.047, I2 = 0.0%) or an increase on the depressive scale: SMD of 0.42 (95% CI 0.03, 0.82, P = 0.036, I2 = 66.6%). Meanwhile, SCH did not associate with depression in older patients (≥60 years old), as defined by the diagnosis of depression: OR of 1.53 (95% CI 0.81, 2.90, P = 0.193, I2 = 71.3%) or an increase on the depressive scale: SMD of 0.03 (95%CI −0.31, 0.37, P = 0.857, I2 = 79.8%). We also found an insignificant difference in the composite endpoint between the L-T4 supplementation group and placebo group in SCH patients. The estimated SMD was 0.26 (95% CI −0.09, 0.62, P = 0.143, I2 = 52.9%). This meta-analysis demonstrates that SCH is not connected to depression. However, sub-group analysis according to age found that SCH is related to depression in younger patients, but not in older patients. Furthermore, we failed to find an effect of L-T4 supplementation treatment for SCH on depression.