BACKGROUND:The mechanisms by which spasticity reductions after botulinum toxin A (BoNT) affect gait in stroke are not well understood. We systematically reviewed the effects of BoNT on spatiotemporal, kinematic, kinetic and electromyographic (EMG) measures during gait. QUESTION:What are the effects of botulinum toxin on gait mechanics in stroke patients? METHODS:Systematic search using PubMed and Web of Science. We considered all studies that reported laboratory-based and instrumented gait measures as primary or secondary outcomes to determine the effects of BoNT on walking performance in stroke populations only. Selected studies were classified and analysed based on the injection sites. RESULTS:A total of 240 articles were identified of which 22 were selected for analysis. Overall, 91% of the studies reported spatiotemporal, 64% kinematics, 23% kinetics, 32% EMG and 23% other gait measures. All but one study found significant effects of BoNT on gait measures using instrumented assessments even when clinical measures (i.e. speed) did not significantly improve. However, the majority of the studies had a high risk of bias. Overall, BoNT improved: a) dorsiflexion during stance, propulsive forces and timing and activity of more proximal musculature with injections in the plantarflexors; b) hip, knee and ankle angles and velocities, coordination and energetic cost with injections in the rectus femoris; c) segmental coordination and energetic cost when several lower limb muscles were injected; and, d) elbow and trunk angles when upper limb muscles were injected. CONCLUSION:Instrumented and laboratory measures of gait improve after BoNT injections in different muscle groups even in the absence of clinical changes.