Two SIMCA models were developed for the classification of acyclic octene isomers, which only form a fraction of a very complex product mixture obtained, for example, from the transformation of 1-butene. The effects of spectral transformation, namely autocorrelation and logarithmic intensity ratios transforms, and (square-root) scaling of the octane isomers mass-spectral data were investigated. Both the spectral-features preprocessing methods and scaling were found to be vital for an adequate development and improvement of the classification models. The best SIMCA models were successfully applied on gas-chromatography-mass spectroscopy (GC-MS) analysis collected from the dimerization of 1-butene over heterogeneous catalysts in the liquid phase.