Abstract Background Haptocorrin carries the major part of the circulating vitamin B 12. The protein is heavily glycosylated and this may have implications for its measurement. Methods We used two different ELISA assays. In one assay, we employed antibodies against native HC and no pre-treatment of samples or calibrators. In the other assay, we used antibodies raised against deglycosylated HC, and deglycosylated the samples and calibrators by treatment with neuraminidase and PNGase prior to analysis. Plasma samples from healthy donors were analysed. Results The ELISA against native HC showed a high detection limit (71 pmol/l) and a poor linearity for serial dilutions of samples. The ELISA against deglycosylated HC showed a detection limit of 1.6 pmol/l, an excellent linearity between 1.6 and 100 pmol/l ( r 2 = 0.99) and an inaccuracy of 5% for concentrations ranging from 250 to 840 pmol/l. The 95% reference interval was 240–680 pmol/l ( n = 148). The concentration of HC showed a strong association to plasma cobalamins ( p < 0.0001). Conclusions An ELISA against native HC does not ensure an equimolar measurement of HC, while this is the case when a glycosylation independent assay is employed. Using this assay, a very strong correlation between total plasma HC and cobalamins in healthy donors is obtained.