Previous brain imaging studies have demonstrated a seasonal difference of serotonin transporter (SERT) binding in the human brain. However, the results were somewhat contradictory. We conducted test–retest study with single photon emission computed tomography (SPECT) with 123I-ADAM as ligand in 28 healthy subjects. Ten of the subjects were studied within 1month, whereas 18 were randomly assigned to be studied over a period of up to 1year. The primary measure was the specific uptake ratio (SUR). Regions of interest included the midbrain, thalamus, putamen and caudate. The intra-class correlation coefficient (ICC) was 0.52–0.94 across different brain regions over 1month, whereas the ICC was -0.24–0.63 over a 1-year period. The 1-month variability ranged from 6.5±5.1% to 12.5±10.6% across different brain regions, and the 1-year variability ranged from 16.5±9.6% to 41.9±35.5%. The Kruskal–Wallis test revealed a significant difference of variability across months. The Wilcoxon Signed Ranks Test showed the SUR between test-retest scans was of borderline significance. Curve fitting, using a 4th degree polynomial model, revealed a significant circadian correlation between the variability and interval of test-retest measurements. Our findings demonstrate the test–retest reproducibility of 123I-ADAM in different time periods and suggest that circadian variation of SERT levels in the human brain might exist.