Ethyl glucuronide (EtG) is a promising new biological state marker of recent alcohol consumption that detects alcohol use reliably over a definite time period. Other currently available markers lack acceptable sensitivity and specificity. Our aim is to elucidate under naturalistic conditions the time course of EtG excretion in urine following alcohol consumption and to show how this can be utilized for monitoring and prognosis in patients. There are no other existing data on this issue to date. One hundred and thirty-eight urine samples from 28 male alcohol withdrawal patients were drawn every 3-24 hours for up to 94 hours after hospitalization. Breath ethanol concentration (mean) at hospitalization was 900 mg/L. Patient age in years was 40.3 (mean). Determination of urine EtG was performed by gas chromatography/mass spectrometry (GC/MS) with deuterium-labelled EtG as an internal standard. The strongest correlations (p<0.01) were found between EtG determinations in the different patient when breath ethanol concentrations (BEC) were 0 and 48 hours after BEC=0 (r=0.747), EtG 24 and 48 hours after BEC=0 (r=0.872), and in the time frame of detection (hours) of EtG and EtG 48 hours after BEC=0 (r=0.762). No significant correlation was found (Mann-Whitney test) between EtG concentrations in urine at different time points between the groups of patients with (a) 1 or less-2, (b) 3-4 or more previous hospitalizations, (c) a history of seizures (yes/no) or (d) an age above or below the median (40.5). EtG excretion in urine is not random, but seems rather to follow a kinetic profile. Furthermore our preliminary data indicate, that there is no significant difference for EtG concentration in urine when correlated to group variables such as age, seizures and hospitalizations.