Tempo is a key determinant of the motivational effects of music during exercise and has been the focus of numerous empirical studies (e.g., Karageorghis & Jones, 2014). The present study sought to address the limitations of previous related work and revisit the relationship between exercise intensity and music-tempo preference using unfamiliar, non-lyrical music (to isolate the tempo manipulation). A within-within experimental design was employed to test hypotheses pertaining to the non-linear relationship and associated psychological outcomes (e.g., core affect and state attention). Twenty-four participants (Mage = 20.6 years, SD = 0.92 years) exercised at five intensities (10% of peak V̇O2 below ventilatory threshold [VT]; 5% of peak V̇O2 below VT, at VT, midway between VT and the respiratory compensation point [RCP], and at RCP) during which they were administered music tracks at four tempi (90 bpm, 110 bpm, 130 bpm and 150 bpm) and a no-music control. A music liking item, measures of core affect (valence and arousal), attentional focus and perceived exertion were recorded during the exercise bouts. Results indicated that unlike previous findings with familiar, lyrical music, there was no discernible relationship between exercise intensity and preference for music tempo. The most positive psychological outcomes were associated with fast-tempo music. In accord with previous findings, slow-tempo music attracted low liking scores and the least desirable psychological outcomes at every exercise intensity. The present findings have implications for the use of unfamiliar, non-lyrical music during exercise. Specifically, that such music should be ∼10 bpm faster than familiar, lyrical music.