The knee adduction moment is consistently used as a surrogate measure of medial compartment loading. Foot orthoses are designed to reduce knee adduction moment via lateral wedging. The 'dose' of wedging required to optimally unload the affected compartment is unknown and variable between individuals. This study explores a personalised approach via three-dimensional printed foot orthotics to assess the biomechanical response when two design variables are altered: orthotic length and lateral wedging. Foot orthoses were created for 10 individuals with symptomatic medial knee osteoarthritis and 10 controls. Computer-aided design software was used to design four full and four three-quarter-length foot orthoses per participant each with lateral posting of 0° 'neutral', 5° rearfoot, 10° rearfoot and 5° forefoot/10° rearfoot. Three-dimensional printers were used to manufacture all foot orthoses. Three-dimensional gait analyses were performed and selected knee kinetics were analysed: first peak knee adduction moment, second peak knee adduction moment, first knee flexion moment and knee adduction moment impulse. Full-length foot orthoses provided greater reductions in first peak knee adduction moment (p = 0.038), second peak knee adduction moment (p = 0.018) and knee adduction moment impulse (p = 0.022) compared to three-quarter-length foot orthoses. Dose effect of lateral wedging was found for first peak knee adduction moment (p < 0.001), second peak knee adduction moment (p < 0.001) and knee adduction moment impulse (p < 0.001) indicating greater unloading for higher wedging angles. Significant interaction effects were found for foot orthosis length and participant group in second peak knee adduction moment (p = 0.028) and knee adduction moment impulse (p = 0.036). Significant interaction effects were found between orthotic length and wedging condition for second peak knee adduction moment (p = 0.002). No significant changes in first knee flexion moment were found. Individual heterogeneous responses to foot orthosis conditions were observed for first peak knee adduction moment, second peak knee adduction moment and knee adduction moment impulse. Biomechanical response is highly variable with personalised foot orthoses. Findings indicate that the tailoring of a personalised intervention could provide an additional benefit over standard interventions and that a three-dimensional printing approach to foot orthosis manufacturing is a viable alternative to the standard methods.