Laboratory and field studies were performed at the River Road Laboratories of Environment Canada, to evaluate the effect of storage containers, tubing material and pumping method on the integrity of groundwater samples contaminated with volatile organics. The work was aimed at establishing a sampling protocol which can be easily adapted to a variety of field situations and which preserves sample integrity at every stage from collection to analysis. In the laboratory, storability studies were conducted on prepared samples containing a suite of 15 volatile aliphatic, halogenated aliphatic and aromatic compounds, in both glass vials and in Tedlar bags developed during the project. Glass was found to be the preferred container for the storability of samples. Leaching studies were carried out on both Teflon and polyethylene tubing (100 ft., 0.2″ i.d.); neither material was found to substantially contaminate organic-free deionized water. Sorption studies conducted on the same tubing sample 9 indicated that polyethylene sorbs the organics to a substantially greater degree than Teflon. However, representative samples can be obtained through both tubing materials provided that a flushing procedure is followed prior to sample collection. In the field, tests were carried out to determine if sampling into glass vials (significant sample/air contact during collection) introduces any bias to prepared standards. The results suggest that a closed sampling system is preferred. However, all vial samples were collected without analytically significant losses of volatiles. Tubing performance was tested at specially designed piezometers constructed so that groundwater samples could be collected from the same place through either polyethylene or Teflon tubing. It was determined that quantitatively similar samples can be collected through both types of tubing. A comparative study between peristaltic and positive displacement pumping methods was undertaken. The Teflon bladder pump (no air contact) was judged superior overall. Peristaltic suction did not appear to bias samples substantially, however, the condition of the piezometer did have a significant effect on sample representivity.