Nearly one in five adults in the United States suffer from mental illness. Emotion regulation techniques are commonly used to assist individuals with mood and anxiety disorders, two of the common types of mental illness. Wearable devices that deliver discreet haptic cues may be useful tools to assist with managing mental illness by allowing an individual to receive biofeedback regarding their mental state. This thesis reports on the design of affective haptic cues, based on music components, that are intended to elicit predictable emotional responses when displayed to the user in a wearable haptic bracelet. The music components of mode, loudness, and tempo were mapped to vibrotactile equivalents using the software tool Syntacts. Two experiments were conducted to evaluate user perception of and the affective response that the cues generated in users. Experiment 1 quantified the distinguishability of these music-based haptic cues. We investigated whether cues based on major and minor chords were distinguishable from one another. The results did not support the hypothesis that major and minor chord based haptic cues would be reliably identifiable. In Experiment 2, we investigated whether the duration and amplitude of a haptic cue affected the emotional response of the participant. The results were supportive of the hypothesis that the duration and amplitude affect the arousal rating of a cue, but there was no effect on valence ratings. The experimental results identify both the challenges and opportunities in using affective haptic cues displayed to users in a wearable device as a tool for emotion regulation.