Purpose: The StartReact effect, whereby movements are elicited by loud, startling acoustic stimuli (SAS), allows the evaluation of movements when initiated through involuntary circuitry, before auditory feedback. When StartReact is applied during poststroke upper extremity movements, individuals exhibit increased muscle recruitment, reaction times, and reaching distances. StartReact releases unimpaired speech with similar increases in muscle recruitment and reaction time. However, as poststroke communication disorders have divergent neural circuitry from upper extremity tasks, it is unclear if StartReact will enhance speech poststroke. Our objective is to determine if (a) StartReact is present in individuals with poststroke aphasia and apraxia and (b) SAS exposure enhances speech intelligibility. Method: We remotely delivered startling, 105-dB white noise bursts (SAS) and quiet, non-SAS cues to 15 individuals with poststroke aphasia and apraxia during repetition of six words. We evaluated average word intensity, pitch, pitch trajectories, vowel formants F1 and F2 (first and second formants), phonemic error rate, and percent incidence of each SAS versus non-SAS-elicited phoneme produced under each cue type. Results: For SAS trials compared to non-SAS, speech intensity increased ([delta] + 0.6 dB), speech pitch increased ([delta] + 22.7 Hz), and formants (F1 and F2) changed, resulting in a smaller vowel space after SAS. SAS affected pitch trajectories for some, but not all, words. Non-SAS trials had more stops ([delta] + 4.7 utterances) while SAS trials had more sustained phonemes (fricatives, glides, affricates, liquids; [delta] + 5.4 utterances). SAS trials had fewer distortion errors but no change in substitution errors or overall error rate compared to non-SAS trials. Conclusions: We show that stroke-impaired speech is susceptible to StartReact, evidenced by decreased intelligibility due to altered formants, pitch trajectories, and articulation, including increased incidence of sounds that could not be produced without SAS. Future studies should examine the impact of SAS on voluntary speech intelligibility and clinical measures of aphasia and apraxia.