During a force-matched bilateral task, when pain is induced in one limb, a shift of load to the non-painful leg is classically observed. This study aimed to test the hypothesis that this adaptation to pain depends on the mechanical efficiency of the non-painful leg. We studied a bilateral plantarflexion task that allowed flexibility in the relative force produced with each leg, but constrained the sum of forces from both legs to match a target. We manipulated the mechanical efficiency of the non-painful leg by imposing scaling factors: 1, 0.75, or 0.25 to decrease mechanical efficiency (Decreased efficiency experiment: 18 participants); and 1, 1.33 or 4 to increase mechanical efficiency (Increased efficiency experiment: 17 participants). Participants performed multiple sets of three submaximal bilateral isometric plantarflexions with each scaling factor during two conditions (Baseline and Pain). Pain was induced by injection of hypertonic saline into the soleus. Force was equally distributed between legs during the Baseline contractions (laterality index was close to 1; Decreased efficiency experiment: 1.16±0.33; Increased efficiency experiment: 1.11±0.32), with no significant effect of Scaling factor. The laterality index was affected by Pain such that the painful leg contributed less than the non-painful leg to the total force (Decreased efficiency experiment: 0.90±0.41, P<0.001; Increased efficiency experiment: 0.75±0.32, P<0.001), regardless of the efficiency (scaling factor) of the non-painful leg. When compared to the force produced during Baseline of the corresponding scaling condition, a decrease in force produced by the painful leg was observed for all conditions, except for scaling 0.25. This decrease in force was correlated with a decrease in drive to the soleus muscle. These data highlight that regardless of the overall mechanical cost, the nervous system appears to prefer to alter force sharing between limbs such that force produced by the painful leg is reduced relative to the non-painful leg.