ABSTRACTThe StartReact test, increasingly popular for assessing cortico‐reticular functioning, is a valid method to influence the firing of reticulospinal tract neurons noninvasively. However, there remains limited evidence on how different stimuli employed in the StartReact test impact motor output in humans. The present study tested elbow flexor responses of 33 adults (aged 26–48 years) to visual stimuli only (LED light), audio‐visual (80 dB) stimuli, and startle‐inducing audio‐visual (120 dB) stimuli sitting with the arm supinated in an electromechanical dynamometer. Surface electromyogram (EMG) recorded muscle activity from the right biceps brachii muscle. Participants were presented with 20 stimuli for each of the three conditions in pseudorandom order with interstimulus intervals of ~8 s. Reaction times were calculated from the stimulus trigger to the initial rise in the EMG signal above 7 × SD from baseline. Rate of torque development (RTD) and EMG signals were recorded throughout and analyzed over their initial 50 ms and 100 ms time‐windows. Reaction times were reduced from visual (169 ± 23) to audio‐visual (140 ± 23) and further reduced to startle‐inducing audio‐visual stimuli (108 ± 19, p < 0.001). While RTD and EMG were consistently greatest following startle‐inducing stimuli (p < 0.001), they were also enhanced following all audio‐visual stimuli over 100 ms (p < 0.05). It appears that startle‐inducing audio‐visual stimuli result in shorter reaction times, increased RTD, and enhanced muscle activity within the initial 50 ms, likely from subcortical upregulation. However, the 100 ms time‐window suggests cortical upregulation following all audio‐visual stimuli considering the longer transmission times.