Atherosclerosis is a complicated and multifactorial disease, induced not only by genotype, but also, even more importantly, by environmental factors. Study on the metabolic perturbation of endogenous compounds may offer deeper insight into development of atherosclerosis. Gas chromatography/mass spectrometry (GC/MS)-based metabonomics was used to profile a metabolic fingerprint of serum obtained from hamsters with induced cholesterol. The deconvoluted GC/MS data were processed by multivariate statistical analysis tools, such as principal component analysis (PCA) and partial least squares projection to latent structure and discriminant analysis (PLS-DA). For the first time we showed a time-dependent development of the model animal from normal to hypercholesterolaemia, and further to early atherosclerosis. Twenty-one compounds were identified as markers involved in the development to atherosclerosis. Identification of the compounds suggests that amino acid metabolism and fatty acid oxidation are significantly perturbed following cholesterol overloading. The data provide novel information to approach the pathophysiological processes of the hypercholesterolaemia and atherosclerosis disease continuum.