The use of activity-dependent interventions has shown some success in promoting recovery of upper limb function in chronic stroke patients. This study compared the neurophysiological and behavioural changes associated with two such rehabilitation protocols: unilateral and bilateral movement training. Twelve chronic stroke patients were randomly assigned to the two training protocols involving six daily practice sessions. Each session consisted of 50 trials of a dowel placement task performed either with both impaired and unimpaired arm moving synchronously (bilateral training group) or with only the impaired arm moving (unilateral training). Kinematic measurements of upper limb movements were made in four unilateral test trials performed prior to and following each practice session. Functional assessments of the impaired upper limb and neurophysiological assessments, using transcranial magnetic stimulation (TMS), of the affected and nonaffected cortical hemispheres were made prior to and following the intervention sessions. Individuals receiving bilateral training showed a reduction in movement time of the impaired limb and increased upper limb functional ability compared to individuals receiving unilateral training. In some patients changes to upper limb function were associated with changes to the cortical representation of a target muscle in the non-affected hemisphere. Overall, these findings suggest that a short-term bilateral training intervention may be effective in facilitating upper limb motor function in chronic stroke patients. (c) 2006 Elsevier B.V. All rights reserved.