AbstractBinocular rivalry, a visual perception phenomenon where two or more percepts alternate every few seconds when distinct stimuli are presented to the two eyes, has been reported as a biomarker in several psychiatric disorders. It is unclear whether abnormalities of binocular rivalry in depression could occur when emotional rivaling stimuli are used, and if so, whether an emotional binocular rivalry test could provide a trait‐dependent or state‐dependent biomarker. In the current study, 34 individuals with subthreshold depression and 31 non‐depressed individuals performed a binocular rivalry task associated with implicit emotional processing. Participants were required to report their perceived orientations of the rival gratings in the foreground and to neglect emotional face stimuli in the background. The participants were retested after an approximately 4‐month time interval. Compared to the non‐depressed group, the subthreshold depression group showed significantly longer perceptual dominance durations of the grating with emotional faces as the background (i.e., implicit emotional dominance) at the initial assessment. However, the abnormality was not found at the follow‐up assessment. More importantly, we found smaller changes in depressive severity at the follow‐up assessment for individuals displaying longer emotional dominance at the initial assessment than for individuals with weaker emotional dominance. The current emotional binocular rivalry test may provide an objective, state‐dependent biomarker for distinguishing individuals with subthreshold depression from non‐depressed individuals.