The acoustical properties of a concert room tend to affect performers’ decisions, especially regarding tempo and agogics. Consequently, the study of the relationship between concert hall acoustics and the musical performance is of great interest to musical performers, and potentially to architects and acousticians as well. A pilot study was devised, enabled by a unique concert hall with mechanically variable acoustics. A concert pianist performed an identical program of two pieces at four trials throughout the same day in the presence of an audience of experienced musicians-researchers, each trial conducted under a distinctive acoustic condition. The trials were recorded for later analysis. The live performances as well as the recordings were assessed individually by the pianist himself and the members of the expert audience. The results showed clear as well as subtle differences between the different performances. The pilot study was followed by a two-year, still ongoing, research project, in which further experimental series of performances have taken place or are underway, using various chamber music constellations, as well as solo flute, organ and choir. In this project, the music performed has included 3–4 pieces from different periods and in different styles. The impression of the live performances from the performer’s own perspective, and also the professional listeners’, has often differed from the experience in listening to the recordings: what was felt during the performance as an ideal live acoustic was often not judged as optimal in the later analysis, especially in terms of agogics and tempo. The preliminary results raise fundamental questions about tempo treatment and artistic/interpretive decisions and promise to give new insights concerning what actually constitutes ‘good’ acoustics and optimal recording conditions from a musician’s professional perspective.