Perception of self-motion involves the integration of visual and vestibular sensory information. Currently, there is limited research exploring visual-vestibular interactions under weak vestibular stimulation. This study investigates the impact of weak inertial stimulation on visual-vestibular bimodal heading perception. A translational XY-axis motion platform equipped with a 46-inch LCD TV was utilized to generate synchronized visual and inertial stimuli. The heading perception was examined under visual-only, vestibular-only, and bimodal conditions using three levels of inertial stimuli (9 mg, 14 mg, and 19 mg). In each condition, participants were tested at nine angles (±16°, ±9.2°, ±3°, ±1.7°, and 0° where 0° represents forward movement), to discern left-forward or right-forward motion. The heading discrimination threshold (HDT) was derived from participants' rightward response proportions across all angles. Our findings reveal that the HDT under 14 mg bimodal conditions is significantly higher than that under visual-only conditions (with marginal significance in the 9 mg and 19 mg conditions), indicating that the presence of weak vestibular signals might decrease the precision of bimodal heading discrimination. These results contradict the prediction of Bayesian model theory that perception is more precise under bimodal compared with unimodal conditions. The results may be explained by neurological biases during Bayesian integration, the "reduced visual precision" theory, or increased task complexity in bimodal heading discrimination. Further research with larger sample size, extending the study to varied inertial stimuli and visual coherence levels, will be beneficial for clarifying its underlying mechanisms.