Air sampling in a government building was necessary in response to reports of a cancer cluster. SKC (Eighty Four, Pa.) charcoal coconut shell-based sorbent tubes (226-01 lot 120) were recommended for this procedure. A recently purchased supply was present at the University of British Columbia and consequently was used for this particular study. Analysis of the front charcoal section showed the presence of a flame retardant, tri (2-chloroisopropyl) phosphate, which was confirmed by gas liquid chromatography (GLC) and mass spectrometry analysis. In an effort to identify the source of this fire retardant in the building, it became apparent from the analysis done on unknown field blanks that tri (2-chloroisopropyl) phosphate was a contaminant of the sorbent tubes used. Analysis of additional blank tubes identified the foam separators as the most likely source of contamination. Levels of tri (2-chloroisopropyl) phosphate in the front charcoal section ranged from 1.3 to 5.9 micrograms. The foam separator contained between 11.4 and 16.5 micrograms, and the backup charcoal section contained between 14.5 and 24.0 micrograms of tri (2-chloroisopropyl) phosphate. In addition, another flame retardant, tri (1,3 dichloro-2-propyl) phosphate was also found. Because these contaminants have long column retention times in GLC, it may not be apparent that these contaminants are present and consequently are likely to have modified the sorbent characteristics of the activated charcoal. Another batch of sorbent tubes bearing the same catalog number and lot number was purchased from the supplier; no flame retardants were found in this batch.(ABSTRACT TRUNCATED AT 250 WORDS)