The constant volume sampling (CVS) technique, which has been part of the Federal Test Procedure (FTP) for the exhaust emissions testing of light-duty vehicles since the 1972 model year, involves the collection of a sample of exhaust gas after it has been diluted to a constant volume. The FTP specifies a formula for calculating a dilution factor (DF) that is used to correct the emission measurement for the pollutant concentration in the dilution air. Once the DF has been determined, emission measurements made using the CVS technique can be converted to a “raw,” undiluted concentration. This enables a single sampling system to be used to determine either mass emissions or tailpipe concentrations, both of which are required in certain vehicle inspection and maintenance (I/M) programs.Review of the DF calculation procedure specified in the FTP indicates that it is a simplification of a more rigorous calculation needed to most accurately determine the true DF. Although the calculated DF is properly used to correct for contamination of the dilution air, the DF calculation procedure specified in the FTP is valid only when the dilution air is uncontaminated. In addition, the DF calculation procedure contained in the FTP is based on the assumption of stoichiometric combustion. Errors are introduced if the vehicle being tested is running richer than stoichiometric. Finally, the DF calculation assumes that emissions are being measured without removal of any water from the sample. This is a good assumption for the way most analytical systems are designed; however, some emissions analysis systems condense water from the sample under certain test conditions.As used in the FTP, the specified DF calculation technique can introduce errors in the mass emissions calculation, but the magnitude of the potential errors is small relative to the current emission standards. When used to calculate raw exhaust concentrations from dilute samples, the DF calculation procedure specified in the FTP can result in errors that are relatively larger. Using a more rigorous derivation of the DF, the accuracy of raw exhaust concentrations calculated from dilute samples is significantly improved. The improved DF calculation procedure would also improve the accuracy of mass emission measurements, which would be desirable when testing vehicles with near zero emissions.