This article presents the benzodiazepine concentrations in urine samples from participants undergoing alcohol withdrawal in a Phase 4 "Proof of Concept" double-blind, randomized, controlled clinical trial. Chlordiazepoxide was prescribed to all participants "as needed" during the first 2 weeks only of alcohol withdrawal, to prevent serious consequences such as seizures. The trial examined effects of either mifepristone or placebo on the primary trial outcomes, which included cognitive function tests at 3 weeks and 4 weeks after the cessation of drinking. Because benzodiazepines are known to affect memory, urine benzodiazepine concentrations were measured before cognitive testing. Urine samples were collected from participants immediately before each cognitive testing session, and the concentrations of unconjugated benzodiazepines (i.e., compounds active at benzodiazepine receptors) were measured by standard assay, using mass spectrometry. The urine benzodiazepine measurements showed clearly that amounts of active benzodiazepine metabolites were present during the third and fourth weeks after the cessation of drinking that were as high as or higher than those seen after therapeutic dosing. The urinary benzodiazepine concentrations demonstrated that residual active benzodiazepine compounds can be present up to 2 weeks after the last ingestion. This could affect the results of cognitive testing in people with alcohol dependence undergoing detoxification.