Background It has been suggested that fetuin-A may be a potential biomarker of cardiometabolic disease. However, few studies have investigated preanalytical factors that might impact the measurement of fetuin-A in the circulation. This pilot study aimed to investigate the preanalytical variables of sample type, timing of sample centrifugation and the impact of freeze–thaw cycles on the concentration of fetuin-A in serum or EDTA-plasma. Methods Blood samples were taken from 19 male or female healthy volunteers, aged 18–70 years, and left at ambient room temperature for 2 h or 48 h. The tubes were then centrifuged, serum and EDTA-plasma separated, and fetuin-A concentrations measured using a commercially available enzyme-linked immunosorbent assay (ELISA). Results There was no significant difference between the concentrations of fetuin-A in EDTA-plasma and serum following separation from whole blood at 2 h postcollection (P = 0.78). The median (interquartile range) concentrations of fetuin-A in EDTA-plasma separated at 2 h and 48 h postcollection were 589 µg/mL (484–703 µg/mL) and 767 µg/mL (687–942 µg/mL), respectively (P < 0.0005). For serum, equivalent concentrations were 606 µg/mL (501–669 µg/mL) at 2 h and 607 µg/mL (564–757 µg/mL) at 48 h postcollection (P = 0.06). Fetuin-A concentrations measured in EDTA-plasma and serum showed no significant change following three freeze–thaw cycles in samples separated at 2 h postcollection (EDTA-plasma P = 0.16; serum P = 0.89). Conclusion This small pilot study has shown that serum is preferable to EDTA-plasma for the measurement of fetuin-A. It has also shown that fetuin-A appears to be as stable after three freeze–thaw cycles as it is after one.