Studies were performed to assess the practicality of using dry VOC-fortified soils as performance evaluation materials (PEMs) and also to compare the effectiveness of EPA's low- and high-level purge-and-trap procedures [SW-846 Method 5030A (EPA 1992)]. Desiccator-dried soils were fortified in large batches with a combination of eight neat VOCs at 1000 ng each VOC/g soil. Three studies were performed on each of three soil types, namely, Hayesville (0.2% organic carbon, 32% clay), Kyle (0.6% organic carbon, 9% clay), and Charleston (3.8% organic carbon, 8% clay). VOC recoveries by the high-level procedure were good to excellent for all soils and compounds (98% mean recovery) and only fair by the low-level procedure (71% mean recovery). The mean relative standard deviation for all soil/compound/procedure combinations in the first study was 11%. Soil and compound properties influenced the VOC recoveries and rates of vapor losses from dry VOC-fortified soils. The poorest recoveries by the low-level procedure occurred on the soil with the highest proportion of organic matter. Rapid volatilization losses of 1,1,1-trichloroethane, benzene, trichloroethene, toluene, and tetrachloroethene were observed from 1-g samples of Kyle soil. In contrast, only trichloroethene was observed to volatilize from the Hayesville soil and no decreases in VOC concentrations were observed on the Charleston soil from 1-g samples that sat open for 20 min. No volatile losses were observed from any of the soils when dry 20-g soil samples were left open for 20 min. The technique of fortifying large batches of dry soils for PEMs as used in these studies provides good precision, but some modifications are suggested to improve the recoveries of compounds with low GC retention times on some of the soils. Also, further studies are needed to evaluate the influence of soil properties on VOC recoveries.