Urinary phenol determinations have traditionally been used to monitor high levels of occupational benzene exposure, but the same technique cannot be used to monitor low-level exposures because of the high background of phenol resulting from its presence in many foods and from metabolism of aromatic amino acids. Thus, new biological indexes for exposure to low levels of benzene are needed. Animal studies indicate that muconic acid is a metabolite of benzene that is excreted in the urine as an increasing fraction of the total benzene metabolities with decreasing dose of benzene. Thus, urinary muconic acid is potentially useful as a monitor for low levels of exposure to benzene. It is also of interest to determine the level of muconic acid in the urine of humans exposed to benzene for comparison with animal data as an aid for use of the animal studies in risk assessments for humans. This report describes the development of a gas chromatography/mass spectrometry assay to detect and quantitate the benzene metabolite, muconic acid, in urine. the internal standard used in the assay, muconic acid-d4, was biosynthesized by F344/N rats administered benzene-d6 by gavage; the muconic acid was isolated from the rat's urine. Muconic acid was measured in experimental urine samples by adding the internal standard, followed by extraction and derivatization. Phenol was also measured in urine after extraction and derivatization. The assays were applied to the urine samples from 14 workers occupationally exposed to benzene and 8 workers with no known benzene exposure. Muconic acid could be detected in all of the urine samples at levels greater than 100 ng/mL. A linear correlation could be fitted to the time-weighted average (TWA) benzene concentrations, as measured by room air monitoring, and muconic acid in urine (R = 0.9). Workers exposed to benzene concentrations of as little as 4.4 ppm over an 8-hr day possessed higher urinary muconic acid concentrations than did any control individual (p < 0.005). In contrast, no significant difference could be detected for urinary phenol between exposed and control workers at these levels. The ratio between muconic acid and phenol in the urine of the workers exposed to benzene (0.38) was similar to that observed in rats (0.31), but not mice (0.62), exposed to similar amounts of benzene. The assay for muconic acid may potentially be useful as a biological exposure index for benzene at levels approaching the current American Conference of Governmental Industrial Hygienists' threshold limit value (TLV) of 1 ppm; however, improved knowledge of the excretion of muconic acid as a function of both dose and time after exposure will need to be made.