Abstract
Summary Twenty retail samples of honey purchased in Sydney, Australia, were analysed for HMF content using both the Codex Alimentarius Commission method (1969) and HPLC. No significant differences in reproducibility or levels found were observed between the methods. Furfural, which was detected in half the samples analysed by HPLC, was shown to cause overestimates of HMF by the Codex method in samples with known amounts of furfural added. Greater differences between the two methods were observed using heated honeys, with the Codex procedure giving higher values. This was not caused by furfural which did not increase in the honeys during heating.

This publication has 5 references indexed in Scilit: