Enzyme-linked immunosorbent assay (ELISA) kits are widely used to quantify corticosterone levels for the assessment of stress in laboratory animals. The aim of this experiment was simply to evaluate if four different and widely used commercial ELISA assays would yield the same or similar values of corticosterone in serum samples taken from laboratory rats after the mild stress of being held for sampling blood from the saphenous vein. Trunk blood was sampled from 32 male Wistar rats 30 minutes after this mild stress exposure and analysed with each of four commercial ELISA kits. Both the Arbor Assays and the DRG-4164 kits were significantly higher than the DRG-5186 and the Enzo kits. There were no significant differences between the DRG-5186 and Enzo kits. Overall the correlations between kits were high. In conclusion, the commercial ELISA kits tested in the present experiment yielded different values of total corticosterone in the same serum samples. The precision in determining true values of the corticosterone level is low for these commercial ELISA kits, although they may be used to determine relative differences within studies.