The case of seven urine samples collected for anti-doping purposes during a cycling stage race with moderately elevated testosterone and epitestosterone ratio (T/E) is reported. The very low probability of having all seven urine samples with such similar elevated T/E ratio (from 3.2 to 4.7) was very suspicious. Different pattern classification tools were tested to categorize the most similar steroid profiles, but none of the models enabled a clear classification of the different urine samples. Subsequently, genetic profiling of all urine samples was performed and demonstrated that three of the seven samples were collected from the same cyclist. Finally, the International Federation confirmed DNA profiling results. This suggests that urinary steroid data using several methodologies are not appropriate for identification purposes and to an extent not unique to individuals.