This paper studies the effect of human perception of visual and audio settings in an urban environment on annoyance. Video clips were projected onto a window panel of a living room to simulate neighborhood views containing different percentages of sea, mountain greenery, and road. These video clips were combined with audio stimuli corresponding to the congruent traffic and sea sounds. 246 participants were presented with 11 audio-visual stimuli and requested to respond to questions after the presentation. The collected responses were used to formulate a multivariate ordered logit model to predict the probability of evoking a high annoyance response. The findings revealed that views embracing mountain greenery close-by could enhance annoyance, which is contrary to other findings that greenery could always moderate noise annoyance. In addition, a 60% sea view was found to be able to yield 1 dB equivalent reduction in total sound pressure level. The trade-off was comparable to that achieved by having sea sound at a level 5 dB higher than road traffic noise. Exposure to road traffic noise level being 3 dB higher than sea sound level (i.e., signal-to-noise ratio = -3) together with a 60% sea view could provide an additional 1.5 dB equivalent reduction.