Several studies have shown that differences in the placement of a headphone over a listener’s ears could result in large differences in the measurements of the related transfer function (HPTF). Nevertheless, because of – at least – the non-uniform frequency resolution of human hearing system, large HPTF variations at some frequencies do not necessary imply audible consequences, which were not evaluated by past studies. The present study aims at evaluating the audibility of spectral modifications introduced by slight but realistic changes in the headphone placement over a listener’s ears. Recordings were performed by placing/replacing a headphone on a dummy head. Various headphone models were realistically placed eight times each on the artificial head. Music excerpts and pink noise then were played back over the headphones and recorded with microphones located at the entrance of the blocked ear canal. These recordings were then presented to expert and naïve listeners over a single test headphone. The subjects had to discriminate between stimuli standing for different headphone placements using a 3I3AFC procedure. With the exception of the naïve listeners about one given music excerpt only, subjects were always able to discriminate between the stimuli with respect to their corresponding headphone placement. This indicates that consecutive realistic headphone placements may result in audible differences for the listener. Such a result could raise several issues about the use of headphones for psychoacoustic experiments, especially for multi-session tests.