A study of the loss of analyte from gas sampling bags was conducted. Five types of sampling bags were investigated, including five layer aluminized bags and Halar, Saran, Tedlar, and Teflon bags. The bags were tested with acetone (67641) 100 parts per million (ppm) in nitrogen and 700ppm in air, 107ppm benzene (71432) in air, 9.5ppm 1,3-butadiene (106990) in nitrogen, 1050ppm 1-butene (106989) in nitrogen, 200ppm methanol (67561) in nitrogen, and 100ppm trichloroethylene (79016) in nitrogen. At selected times between 0 and 24 hours after the bags were filled, the contents of each bag were examined by injecting an aliquot into a gas chromatograph. All bags showed a loss of analyte. The aluminized bags showed a sharp initial loss that tended to level off by 6 to 12 hours. The other bag materials lost analyte continuously, the smallest loss occurring with the Tedlar bags. Bags were also filled with 700ppm acetone and allowed to stand for a week, then the bags were filled with clean air or 107ppm benzene. The rate of disappearance of analyte from the bags was monitored as before. The acetone concentration in the aliquots gradually increased with time, whereas the benzene concentrations decreased. The authors conclude that a problem of sample integrity exists when sampling with bags. In the experiments with acetone plus benzene, a competitive effect between acetone and benzene occurred. For long term samples, preequilibrated aluminized bags seem to be the best choice, whereas for short term sampling, Tedlar bags seem to be the best. A multifactorial experiment designed to evaluate such factors as material of bag construction, humidity, analyte construction, bag size, and bag to bag variation is in progress.