Reducing the knee adduction moment (KAM) is a promising treatment for medial compartment knee osteoarthritis (OA). Although several gait modifications to lower the KAM have been identified, the potential to combine modifications and individual dose-responses remain unknown. This study hypothesized that: (i) there is a general scheme consisting of modifications in trunk sway, step width, walking speed, and foot progression angle that reduces the KAM; (ii) gait modifications can be combined; and (iii) dose-responses differ among individuals. Walking trials with simultaneous modifications in step width, walking speed, progression angle, and trunk sway were analyzed for 10 healthy subjects. Wider step width, slower speed, toeing-in, and increased trunk sway resulted in reduced first KAM peak, whereas wider step width, faster speed, and increased trunk sway reduced the KAM angular impulse. Individual regressions accurately modeled the amplitude of the KAM variables relative to the amplitude of the gait modification variables, while the dose-responses varied strongly among participants. In conclusion, increasing trunk sway, increasing step width, and toeing-in are three gait modifications that could be combined to reduce KAM variables related to knee OA. Results also indicated that some gait modifications reducing the KAM induced changes in the knee flexion moment possibly indicative of an increase in knee loading. Taken together with the different dose-responses among subjects, this study suggested that gait retraining programs should consider this general scheme of modifications with individualization of the modification amplitudes. © 2016 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 34:1547-1556, 2016.