W e report here the formation in vivo of a protein-acetaldehyde adduct (protein-AA) in liver when rats were fed alcohol chronically. This chemically modified protein was demonstrated by electroimmunotransblot technique and with rabbit polyclonal antibodies that recognize acetaldehyde adduct as an epitope (i.e., both anti-hemocyanin-AA IgG and anti-myoglobin-AA IgG). It has a molecular weight of 37,000. It can be detected in the liver of rats fed the alcohol-containing American Institute of Nutrition 1976 liquid diet for only 1 wk. Since the protein profiles of soluble hepatic proteins from alcohol-fed and control rats were identical on SDS-PAGE, the peroxidase-positive band demonstrated by electroimmunotransblot was most likely not a new protein synthesized de novo. Borohydride reduction was not necessary to stabilize this protein-AA. Intraperitoneal injections of ethanol (2 g/kg body wt) at 8-h intervals to rats over a 24-h period did not produce any detectable protein-AA in the liver. Incubation of the liver homogenate from a control liver with acetaldehyde without sodium cyanoborohydride for 4 h also failed to generate any protein-AA. Therefore, the formation of the 37-kD protein-AA in vivo reported here is dependent on chronic alcohol consumption.