<p> In this thesis, hardware and software was developed whereby the feasability of using pattern recognition techniques for gas chromatography could be quantitively explored. Processing methods were developed to overcome ~he distortion inherent in GC spectra. The reproducibility of peak retention times and areas was investigated from the processed spectra. Upon completion of this study, it was possible to amass a large quantity of experimental data for bacterial strains. With such data the exact form of the pattern recognition algorithm could be defined. A possible pattern recognition method is discussed. </p> / Thesis / Master of Engineering (MEngr)