Ambient temperature may affect venous glucose concentration after gluc
ose tolerance tests. We analysed 1030 standardised 75 g tests. Althoug
h mean fasting values did not differ, post-load values did: adjusted m
ean 2 h glucose concentration was 1.03 mmol/L lower at lower (5-14 deg
rees C) than at higher (25-31 degrees C) temperatures (p<0.001). The o
ccurrence of abnormal glucose tolerance doubled on warmer days. The di
agnostic accuracy of the glucose tolerance test showed clinically sign
ificant temperature-associated variation. These variations, if confirm
ed, call for temperature standardisation during glucose tolerance test
ing and/or alternative strategies for use when standardisation is not
feasible.