Two new techniques have been applied to FTIR emission analysis which add significant potential to automotive emission measurement. One of these is the use of the mathematical multivariate analysis which is called the partial least squares method. This spectrum discrimination technique, in combination with high resolution spectrum data, enables superior analysis for heavy-overlapping species in the emission. The other technique is a flow conditioned gas sampling cell which is designed especially for real time emission measurement. The flow in the gas cell has been analyzed with computer simulation and the gas cell has a flow conditioner inside with a 10 meter optical path. Seven seconds of 90 percent gas replacement time can be achieved with this cell. As a result, highly accurate realtime data can be obtained with relatively fast response.In this paper, spectrum factors extracted from overlapping species and quantification simulations are shown using standard gases. The realtime measurement of methanol fueled vehicles with and without catalyzer will be compared. Additionally, we will compare data from the FTIR with data collected by conventional analyzers. The results show the detailed emission pattern of several species which have not been analyzed before.