In our daily life, we often receive stimuli from different modalities. Multi-sensory stimuli can generally enhance our processing of information compared with unimodal stimuli. During the processing of sensory information from different modalities, the processing advantage of visual modality over other sensory modalities is called Visual dominance effect. In the past decades such inter-sensory bias among different modalities has been explored extensively. Directed attention hypothesis believes that attention plays an important role in the visual dominance effect, which appears only when attention is focused on the visual modality. Previous studies have found that visual dominance effect could be modulated by endogenous attention or exogenous attention, and most empirical literature have focused on the impact of endogenous attention upon the visual dominance effect or Colavita effect. In the present study, we investigated the influence of bottom-up attention on the visual dominance effect. Specifically, we were interested in whether the salience of visual or auditory stimuli had an impact on the visual dominance effect. By adopting a spatial task-switching paradigm, a unimodal cue (visual or auditory) was presented before the bimodal targets, in which visual and auditory targets were presented simultaneously. On each trial, visual and auditory targets were presented either in left or right side from the center of the screen, which could be either congruent or incongruent. Participants were required to judge the spatial position of targets according to the initial cue. Participants usually respond faster in congruent trials, which is referred to Cross-modal spatial congruence effects. In Experiment 1, we used the salient auditory stimuli to explore whether the salience of auditory stimuli affected the visual dominance effect. The results of Experiment 1 showed that shorter reaction times (RTs) and higher accuracy rates (ACCs) for visual targets than those for auditory targets, and reaction times were faster when the auditory-visual targets were spatially congruent than when the spatial locations of auditory and visual targets were incongruent. For visual targets, the RT differences between congruent and incongruent trials was smaller than that for auditory targets. Furthermore,the visual dominance effect decreased on the condition of salient auditory stimuli. In Experiment 2, visual stimuli of low salience were used to further investigate the change of visual dominance effect. Consistent with the results in Experiment 1, spatial congruence effect was more significant for auditory targets than for visual targets in Experiment 2 and visual dominance effect decreased on the low salient visual condition. Moreover, robust visual dominance effect was observed in both Experiment 1 and 2. In summary, these results indicate that the salience of auditory and visual stimuli could modulate the size of visual dominance effect, although the visual dominant effect is robust. Our results support that exogenous attention of stimuli impacts on the sensory dominance effect. Moreover, visual dominance effect cannot be fully explained by adaptive attentional mechanism, which could be explained by Biased competition hypothesis in the cross-modal interaction. In the process of audio-visual interaction, visual modality has processing advantages over other modalities. Visual stimuli are more salient and tend to attract our attention in purely bottom-up ways, while the processing of auditory stimuli is inhibited.