Isolation of high-purity albumin from plasma is essential to study albumin kinetics in vivo with tracer techniques. Because of its simplicity ethanol extraction has been repeatedly used for albumin purification. However, it cannot be excluded that this single-step procedure completely prohibits contamination by other proteins, especially those known to be produced at an accelerated rate during the acute phase response. In the present study, we wanted to examine the reliability of ethanol extraction in different clinical conditions and to study the effects of potential impurities on albumin enrichment during stable isotope tracer studies. SDS-PAGE revealed a contaminating protein band at about 25,000 Da in healthy subjects and postoperative patients during the acute phase response, but not in critically ill patients. According to densitometry about 8% of proteins after ethanol extraction were contaminants. To examine potential contaminant effects on tracer enrichment 1-[ 13C]-leucine was given to healthy subjects and postoperative patients. Blood samples were taken after various amounts of time, and albumin enrichments (tracer/tracee ratios) were determined from isotope ratios obtained by mass spectrometry. Irrespective of the magnitude of tracer enrichment, postoperative tracer/tracee ratios were significantly higher (on average +10%) in samples exclusively analysed by ethanol extraction than in samples which had undergone additional electrophoretic purification. No significant effect of the contaminant was seen in healthy subjects. N-terminal protein sequencing revealed contaminants to mainly consist of apolipoprotein A-1. Its physiology and pathophysiology may sufficiently explain its variable effects of albumin enrichment. Our findings suggest that exclusive ethanol extraction is inappropriate for albumin isolation in tracer studies performed during the acute phase response. Ethanol extraction may also not be advisable in all other situations known to be associated with a rise in apolipoprotein A-1 turnover.