A study was conducted to assess the effects of controlling filter face temperatures and two differently sized collection systems on diesel total particulate matter (TPM) and vapor phase hydrocarbon levels from a diesel engine. The results were used to revise sampling protocols so that variability associated with quantitation of polynuclear aromatic hydrocarbons (PAH) is minimized.Particulate soluble organic fraction (SOF) levels (%) were compared 1) for tests where the dilute exhaust filter face temperature was held constant by varying dilution ratio (DR) to account for day to day variations in inlet air temperature to the tunnel and 2) for tests in earlier studies where the DR was held constant and the filter face temperature then varied because of varying tunnel inlet air temperature. Between date variations in %SOF were reduced by about 60% due to holding filter face temperatures constant, compared to holding DR constant.Using 47-mm and 508 x 508-mm filter size dilute exhaust sampling systems, SOF and semi-volatile hydrocarbon (collected on XAD-2 resin and termed XAD-2 resin organic component or XOC) samples were collected from several different engine test runs with the same filter face temperature at each run. Individual SOF values for a given day for either sampling system were not significantly different from each other and there generally was no significant difference between SOF levels obtained from the two sampling systems for samples collected on the same date. There were no significant differences in 508 x 508-mm XOC levels either within or between test dates.Measurements of six biologically active PAH were made from SOF collected on two different dates using the 508 x 508-mm filter system. No significant differences were found in levels of any PAH collected on the two different test dates. With benz[a]anthracene plus chrysene, nitropyrene, and fluoranthene, the variability in detected levels was approximately equal to the variability in SOF values for a given engine run. Levels of the other PAH had greater variability, indicating additional effects of sample handling and quantitation when PAH are present near detection limits. When compared to literature values for PAH from SOF/TPM collected from other diesel engines, these PAH concentrations are of comparable magnitude but generally of less variability.