Ejaz et al. (2018) are to be commended for showing no evidence for a cortical origin of post-stroke mirror movements. Using functional MRI during affected-finger presses in recovering adult-onset stroke patients, they found no consistent relationship between contralesional sensorimotor cortex (cSM1) activation and quantitative indices of mirror movements; specifically, mirror movements were not linked to the presence of cSM1 overactivation, arguing against the classic ‘transcallosal’ mechanism heretofore widely believed to cause mirror movements (Di Pino et al., 2014). We wish to report findings—previously published in abstract form (Calautti, 2008)—that further support the idea that mirror movements are not cortically mediated. We also present data that confirm that mirror movements can involve the affected (i.e. paretic) hand during movement of the unaffected (i.e. non-paretic) hand, also arguing in favour of disruption of a bilaterally-organized system.