To test the suggestion that chlorpropamide-alcohol flushing (CPAF) resembles the disulfiram effect and might be mediated by acetaldehyde, the initial metabolite of alcohol, blood concentrations of acetaldehyde were measured after a drink of alcohol in controls and diabetics positive and negative for CPAF. The CPAF-positive diabetics had significantly greater blood acetaldehyde concentrations after alcohol than the CPAF-negative diabetics both with a single dose of chlorpropamide and after two weeks' chlorpropamide treatment. Concentrations in the CPAF-positive group after chlorpropamide were also significantly greater than after a placebo tablet. There was also a clear separation in the increase in facial temperature after two weeks of chlorpropamide between the CPAF-positive and CPAF-negative groups (although there was some overlap after a single tablet). There was no difference in plasma chlorpropamide or alcohol concentrations between CPAF-positive and CPAF-negative diabetics. These findings show that CPAF is distinct from alcohol flushing and that the acetaldehyde concentration in the blood provides an objective measure of CPAF. The difference between flushing and non-flushing diabetics cannot be accounted for by differences in blood concentrations of chlorpropamide or alcohol.