Music is vulnerable to unconscious bias through visual extramusical factors. Traditional male/female gender bias affects instrument choice and impacts performance evaluation negatively. This study investigates gender bias in assessing harp performance. Harp performers (one male, one female) recorded three musical excerpts, all dubbed with the female performer's audio. Listener-viewers assessed performances on five criteria and were blinded to the study purpose. Follow-up interviews explored awareness of unconscious gender bias. There were no significant differences between ratings for each gender but there were significant interactions between gender and musical excerpt for overall performance quality, musicality and sensitivity. Interviews revealed no explicit gender bias but discovered feminine associations for harp performance resulting in gendered descriptors rather than musical assessments. Listener-viewers demonstrated preconceived gendered expectations of harpists and were susceptible to gender bias in their evaluations of each musical excerpt. Perceptual evaluations can inform and educate future music assessors about unconscious gender bias.