Pathologiessuch as stroke or cerebral palsy are often characterized by excessivemediolateral bending of the trunk. However, it is unclear how mediolateral trunk-sway affects mechanicalwork of more distal joints during gait. We therefore pose the researchquestion: How does increased mediolateraltrunk-sway affect the total system work and that of distal joints during gait? Instead of gait experimentsthat ask subjects to walk with increased mediolateral trunk-sway, optimalcontrol theory is used to find control sequences that impose mediolateral-trunksway on normal gait patterns of 10 healthy subjects. In order to findout how mediolateral trunk-sway affected mechanical output of gait, resultingpower and work metrics are compared to those of normal gait. Optimal controlproblems were solved by transforming the mediolateral trunk-sway induction (TSI)problem into a nonlinear programming problem via direct collocation. The TSIproblem was applied to a 31 DOF rigid body skeletal model with in total 25ideal torque actuators for each subject. Foot ground interaction was simulatedby 4 Hunt-Crossley spheres under each foot.               Exaggeratedtrunk-sway resulted in a total increase of positive work of around 2% (p =0.871). Push-off work (~50-65% of gait cycle) decreased by 40% (p = 0,001)and rebound work (~15-35% of gait cycle) increased by 81% (p = 0,001).Analysis of positive work contribution about individual joints show a massive660% increase in total positive lumbar work (p < 0,001), as well as a35% decrease in positive ankle work (p = 0,01) during push-off.Additionally, ground reaction force impulse during the push-off phase decreased by 37% (p =0,004). Mediolateral trunk-sway gaitwas successfully imposed on normal gait by using optimal control theory. It wasdemonstrated that mediolateral trunk-sway is not an efficient gait pattern,since total work increased. Most of this increase is due to increases in lumbarjoint work as a direct result of the imposed movement pattern. Mediolateraltrunk-sway also resulted in a significant decrease in required ankle work,combined with a small increase in hip work. Additionally, ground-reactionanalysis showed a decrease in impulse measured during the push-off phase as aresult of imposing trunk-sway. These combined factors, could indicate that these joints might beused for forward propulsion of the body within patients that use mediolateraltrunk-sway as an adaptation.