The proportion of hemoglobin A1 was measured in 43 normoglycemic men and women admitted consecutively to an alcohol treatment center and in 41 healthy volunteers. On the average, the size of the fraction was significantly greater in the alcoholic patients than in the healthy subjects. The results of hemoglobin A1c determinations added independent evidence that this difference was not due to formation of any of the known glycosylated hemoglobins despite the fact that the alcohol-related hemoglobin was found to cochromatograph with hemoglobin A1c. The size of the hemoglobin A1 fraction was unrelated to the presence or absence of alcoholic hepatitis, although most patients having abnormally high serum hepatic enzyme activities were also found to have abnormal elevations of the A1 fraction. However, some patients with normal serum enzyme activities were among those having the highest values of the A1 fraction, indicating that the latter was a more sensitive indicator of excessive alcohol intake than were hepatic enzyme activities. The proportion of hemoglobin A1 in a patient followed over a period of 9 months rose to a maximum as the patient continued to drink. In a patient who abstained from drinking while under hospital care, the A1 fraction decreased towards normal with a half-time of 11 days. These observations are interpreted as evidence that alcohol-related hemoglobin is a product of a readily reversible reaction occurring at a maximal rate of 0.038 mM day-1, about 4 times as rapid as is the formation of hemoglobin A1c.