["Addiction, Volume 121, Issue 6, Page 1474-1482, June 2026. ", "\nAbstract\n\nBackground and aim\nAccurately estimating alcohol consumption is crucial for clinical decision‐making and monitoring treatment outcomes. Phosphatidylethanol in blood (B‐PEth), a direct alcohol biomarker, is currently the most reliable indicator of alcohol intake, with a detection window of several weeks; however, the factors influencing the decline of B‐PEth levels remain largely unknown. This study aimed to investigate the decline of B‐PEth levels during abstinence in patients with alcohol use disorder (AUD) undergoing withdrawal treatment.\n\n\nMethod\nA total of 100 patients were recruited in withdrawal treatment and followed during three to four weeks of abstinence. Blood samples were collected at baseline and weekly during abstinence to measure B‐PEth levels of two homologues (16:0/18:1 and 16:0/18:2). Self‐reported alcohol consumption was documented using the Timeline Followback (TLFB) method for 30 days before abstinence and throughout the study period.\n\n\nResults\nB‐PEth elimination followed first‐order kinetics. The mean half‐life was 7.24 days [95% confidence interval (CI) = 6.98–7.53] for 16:0/18:1 and 4.55 days (95% CI = 4.44–4.67) for 16:0/18:2. The rate of decline varied by week, with a longer half‐life observed in week three compared with week one. No statistically significant sex differences were detected. The strongest correlation between B‐PEth levels and self‐reported alcohol consumption was found for data from two weeks prior to abstinence.\n\n\nConclusion\nElimination of phosphatidylethanol in blood (B‐Peth) follows first‐order kinetics, with homologue 16:0/18:1 exhibiting a longer half‐life than 16:0/18:2. The rate of decline is influenced by the week of alcohol abstinence and B‐PEth levels are detectable even in the fourth week of abstinence. The strongest correlation between B‐PEth and self‐reported alcohol consumption is at two weeks prior to abstinence.\n\n"]