Purpose: Rehabilitation professionals typically use motor imagery (MI) or action observation (AO) to increase physical strength for injury prevention and recovery. Here we compared hamstring force gains for MI during AO (AO+MI) against two pure MI training groups. Materials and methods: Over a three-week intervention physically-fit adults imagined Nordic hamstring exercises in both legs simultaneously, and synchronised this with a demonstration of the same action (AO+MI), or purely imagined this action (pure MI), or imagined upper-limb actions (pure MI-control). Eccentric hamstring strength gains were assessed using ANOVAs, and magnitude-based inference (MBI) analyses determined the likelihood of clinical/practical benefits for the interventions. Results: Hamstring strength only increased significantly following AO+MI training. This effect was lateralised to the right leg, potentially reflecting a left-hemispheric dominance in motor simulation. MBIs: The right leg within-group treatment effect size for AO+MI was moderate and likely beneficial (d = 0.36), and only small and possibly beneficial for pure MI (0.23). Relative to pure MI-control, effects were possibly beneficial and moderate for AO+MI (0.72), though small for pure MI (0.39). Conclusions: Since hamstring strength predicts injury prevalence, our findings point to the advantage of combined AO+MI interventions, over and above pure MI, for injury prevention and rehabilitation.