We describe methods for the determination of endogenous ethanol in biological specimens from healthy abstaining subjects. The analytical methods were headspace gas chromatography (GC) for plasma samples and gas chromatography-mass spectometry (GC/MS) with deuterium labelled species 2H3-ethanol and 2H5-ethanol as internal standards for breath analysis. Ethanol in rebreathed air was about 10% higher than in directly analysed end-expired alveolar air. Known volumes of rebreathed air were passed through a liquid-N2 freeze trap and the volatile constituents of breath were concentrated prior to analysis by GC or GC/MS. Besides endogenous ethanol, peaks were seen on the chromatograms for methanol, acetone and acetaldehyde as well as several as yet unidentified substances. The endogenous alcohols ethanol and methanol were confirmed from their mass chromatograms and the GC/MS profile also indicated the presence of endogenous propan-1-ol. The concentration of endogenous ethanol in plasma showed wide inter-subject variations ranging from below detection limits to 1.6 micrograms/ml (34.8 mumol/l) and with mean +/- SD of 0.39 +/- 0.45 micrograms/ml (8.5 +/- 9.8 mumol/l). We aim to characterise further the role of endogenous ethanol with the main focus on dynamic aspects such as the rate of formation and turnover.