Facilitating trunk stability is one of the most important objectives in human balance control. This is especially evident in individuals with spinal cord injury (SCI) who are typically not able to control seated balance on their own. As a consequence of their injury, they often suffer under a reduction of strength and control of their trunk muscles and, hence, under reduced independence during daily activities. One of the top priorities for individuals with SCI is therefore to improve their trunk control and, consequently, quality of life. To enhance existing therapies and/or develop bio-inspired assistive technologies that can facilitate dynamic trunk stability in these individuals, a more comprehensive, quantitative understanding of the neuromechanical mechanisms of dynamic sitting control in non-disabled individuals is needed. The objective of this study was therefore twofold:​ (1) to quantify the effect of varying levels of seat instability as well as of visual information elimination on postural efficiency during continuous, multi-directional perturbations using a wobble board paradigm;​ and (2) to quantify the temporal and spatial relationship between muscle activity and wobble board motion during the perturbations. 15 non-disabled individuals were asked to sit on a wobble board inducing continuous, multi-directional perturbations and maintain an upright sitting posture as closely as possible. Five different hemispheres with decreasing diameter were attached to the bottom of the wobble board to induce five different levels of seat instability. Sitting tasks were performed with eyes open and eyes closed. A motion capture system was used to collect trunk and pelvis kinematics as well as those of the wobble board. The activity of fourteen major superficial trunk and upper leg muscles was recorded via an electromyography system. Wobble board kinematics and muscle electromyography were then used to characterize trunk control and stability during dynamic sitting. In a first step, posturographic analyses in time and frequency domain were performed to assess postural proficiency. In a second step, cross-correlation analysis was applied to identify temporal and spatial determinants of muscle activation and, hence, reactive trunk control for the wobble board task. For the posturographic analyses, our findings revealed that time-domain measures were generally increased and frequency-domain measures generally decreased when task difficulty was increased. Similarly, time-domain measures were generally increased and frequency-domain measures generally decreased when visual information was eliminated. For the cross-correlation analysis, our findings indicate the existence of a relation between phasic muscle activation/deactivation and wobble board motion, which increased in intensity with higher levels of seat instability, irrespective of eye condition. Spatial features revealed that the rectus abdominis, erector spinae, biceps femoris, and rectus femoris muscles were correlated with anterior-posterior wobble board displacement, whereas the external oblique muscles were correlated with medial-lateral wobble board displacement. Moreover, temporal features revealed that, regardless of base, eye condition, and wobble board displacement direction, muscle activation/deactivation preceded the wobble board displacement. On the one hand, the posturographic findings suggest that, by increasing seat instability or eliminating vision, the control effort increases and the degree of stability decreases. On the other hand, the cross-correlation results indicate that the dynamic balancing task is accomplished with significant contributions from active control mechanisms that originate from the central nervous system (CNS). More specifically, the spatial characterization suggests that the CNS modulates the phasic muscle activity levels to break the upcoming wobble board motion. For sagittal plane motion, this is done by increasing the effective stiffness between the human body and the wobble board. For frontal plane motion, further work is needed to confirm or dispute the use of such CNS-based stiffness control strategy. The temporal characterization suggests that the CNS takes advantage of the velocity information of the body and/or wobble board to generate the required motor command in advance of an imminent displacement. These interpretations demonstrate that the performed work has made significant contributions to our fundamental understanding of human balance control in general and of wobble board stabilization more specifically. The gained knowledge may be beneficial for enhancing existing therapies and quantitative assessments, but also for developing bio-inspired assistive technologies that can facilitate trunk stability in individuals with SCI.