Music is a powerful tool to convey emotions, and composers use a variety of compositional tools to target specific emotional responses. Here, we investigated how chord gender (major, minor), average pitch height, and chord complexity interact with musical context when evoking emotional responses in listeners. Forty participants listened to musical chord sequences and rated the sequences' valence and arousal. 15-23 chords long sequences were generated by drawing chords from one of 8 chord-pools which were designed to differ in chord gender (major, minor), chord complexity (triads, extended chords), or musical context ("function", "non-functional"). We defined the "functional" context as chord-pools that comprised the I, IV, V and VI chords from the diatonic scale. We defined the "non-functional" context as chord pools that consisted of chords which are parallel shifted by a minor third, a tritone, and a sixth, thus avoiding unaltered diatonic-functional relationships between any of the chords within a pool. Bayesian mixed effects models show a strong impact of sequence gender, average pitch height, complexity, and functionality on participants perceived valence, with functionality and complexity being significantly stronger predictors than chord gender. The responses for arousal showed further strong interactions between the different predictors, with sometimes even changing directionalities of the effects as a function of musical context. The present results further support that coining major happy and minor sad may be a misleading simplification of the rich emotional interactions of musical features, and shed some light on musical functionality as a mediator in psychoacoustic emotional effects. Felix Dobrowohl | Senior Technical Officer - MARCS Institute Western Sydney University - Bldg 5.G.57 Bankstown Campus P: +61 02 9772 6723 E: F.Dobrowohl@westernsydney.edu.au<mailto:F.Dobrowohl@westernsydney.edu.au>