A data set consisting of fourteen isotopic ratios or quantities derived from such ratios for samples of acetylsalicylic acid (aspirin), commercialized by various pharmaceutical companies from different countries, was analyzed. The goal of the data analysis was to explore whether results can be linked to geographical origin or other features such as different manufacturing processes, of the samples. The methods of data analysis used were principal component analysis (PCA), robust principal component analysis (RPCA), projection pursuit (PP) and multiple factor analysis (MFA). The results do not seem to depend on geographic origin, except for some samples from India. They do depend on the pharmaceutical companies. Moreover, it seems that the samples from certain pharmaceutical companies form clusters of similar samples, suggesting that there is some common feature between those pharmaceutical companies. Variable selection performed by means of MFA showed that the number of variables can be reduced to five without loss of information. (c) 2005 Elsevier B.V. All rights reserved.