Control of limb movements may be impaired after stroke due to the loss of connectivity between the cerebral cortex and spinal cord. A notion to improve motor function in stroke survivors is to use alternate motor fibers, such as the reticulospinal tract (RST), which originate from the brainstem and terminate at different levels of spinal cord. One way of targeting the RST is to use a "StartReact" protocol to foster premature release of a preplanned movement in response to a startling stimulus. Our aim was to find support for the preservation of such StartReact effect in stroke survivors. We conducted a systematic review with meta-analysis of literature published in English up to September 2020, to explore differences in motor responses to startling stimuli in StartReact effects. Protocol of the study was registered (PROSPERO Registration No. CRD42020191581). PubMed, Google Scholar, Web of Science, PsycINFO, and Science Direct were searched for relevant literature. The meta-analysis contained six studies involving a total of 151 stroke and healthy participants. Muscle onset latency data were extracted from the qualifying studies and compared using RevMan. StartReact effect was present in both stroke and healthy groups, represented by shortened muscle onset latency when startling stimulus was present. There was considerable heterogeneity of the outcome measures, which was attributed to the range of motor impairments among stroke survivors and methodologies used. Our findings support the notion of preservation of preprogramming ability and suitability of RST and StartReact effect for motor rehabilitation following stroke.