The chlorpropamide-alcohol flush (CPAF) phenomenon was quantitatively related to blood levels of acetaldehyde and chlorpropamide in 105 Type II diabetics, of whom 74 had not previously taken the drug and 31 were on chronic treatment. Standardized skin temperature recordings were made with a sensitive probe. Plasma ethanol and acetaldehyde concentrations were determined by gas chromatography, and those of chlorpropamide by high-pressure liquid chromatography. There were significant positive correlations between plasma acetaldehyde and the skin temperature increase, between plasma chlorpropamide and plasma acetaldehyde, and between plasma chlorpropamide and the skin temperature increase. CPAF-positive patients became CPAF-negative and vice versa following reduction and increase, respectively, in the dose of chlorpropamide. Thus, the CPAF reaction is a consequence of chlorpropamide inhibition of the oxidation of ethanol-generated acetaldehyde, and it appears that the plasma concentration of chlorpropamide is critical. It remains an open question whether the CPAF test has any prognostic value.