Abstract Distinguishing between several cigarette types simultaneously by glass capillary gas chromatography requires the use of multivariate statistics for data reduction, pattern extraction and ranking the importance of the chromatographic peaks. These techniques are here applied to the gas phase of the cigarette smoke. Ten cigarettes were studied, both cased and uncased of 100% bright, 100% burley, 100% oriental tobacco and blends of 33%/33%/33% and 60%/30%/10%, respectively. At least five chromatographic profiles were obtained for each different cigarette, giving a total of 51 chromatograms to serve as the data base. From each chromatogram, containing about 100 peaks, a subset of 29 peaks was selected manually. The data were evaluated by using discriminant analysis and factor analysis; the former technique produced a satisfactory separation of cigarette types into distinct groups. Results from the factor technique were successful in providing discrimination based on three factors.