||The goal of this project was to develop a sample preparation procedure for the gas chromatographic analysis of priority pollutants in water using the technique known as microextraction. A microextraction procedure is one in which an aqueous sample is extracted with a much smaller volume of an immiscible organic solvent, thereby concentrating the analytes in the process of extracting them. Three groups of priority pollutants, the phenols, the phthalate esters and the nitroaromatics were used as model compounds. The procedure currently recommended by the U.S. Environmental Protection Agency (EPA) for the analysis of these compounds involves a large scale extraction followed by evaporation of the extraction solvent. It was felt that microextraction would eliminate some of the difficulties of the EPA procedure such as lengthy sample preparation time, potential loss of volatile analytes and interference from impurities in the solvent which are concentrated during evaporation. The primary criteria used to evaluate the procedure were the recovery of analytes and the precision. A number of factors were investigated to determine their effect on these criteria. These factors included the choice of extraction solvent, the ratio of sample to solvent volume, the extraction time and technique, the effect of inorganic salts and the effect of sample concentration. In the final procedure, a 100.0 mL sample is extracted with 3.0 mL of methylene chloride. Prior to extraction, 30.0 g of sodium chloride are dissolved in the sample. This step greatly improves the recovery of many of the compounds investigated. The extraction is carried out by vigorously shaking the samples by hand for five minutes. Mechanical mixing of the two phases was found to be inefficient, resulting in relatively long equilibration times. The precision of the procedure is excellent; the relative standard deviation in the recovery of each of the compounds investigated was less than 2%. The percent recovery of most of the compounds was 90% or better and the detection limits for most were less than 0.5 parts-per-million.