OBJECTIVE: The relationship between resting-state heart rate variability (HRV) and somatic symptom disorder (SSD) has been explored in several studies. We planned to examine HRV during the performance of psychological tasks, and assumed that it could be applied for the diagnosis of SSD. METHODS: Fifty-three patients with SSD and 52 healthy controls were recruited. Demographic and psychological data were collected. Subjects\' HRV was measured over five 5-minute blocks during which they viewed different scripts (resting state; health anxiety; somatic distress; depression; neutral topic). After each block they completed a self-assessment manikin. Generalized estimated equation analysis was used to analyze the impact of mental scripts, SSD and sex on physiological and psychological indicators. RESULTS: In men there was a script*SSD interaction concerning high-frequency power (HF) involving the neutral script; SSD men showed significantly higher HF than healthy men when viewing neutral script. In women there were script*SSD interactions with respect to low-frequency power (LF) and HF involving the somatic distress script and a script*SSD interaction with respect to LF and emotional valence involving the neutral script; SSD women revealed significantly lower LF and HF than healthy women when viewing somatic distress script, they also showed significantly lower LF and higher valence than healthy ones during neutral script. The somatic distress script*SSD interaction still influenced LF and HF in women after adjusting for potential confounders. CONCLUSION: In women cue-specific HRV is a better method of differentiating people with SSD from healthy individuals than resting-state HRV. TRIAL REGISTRATION INFORMATION: The Research Ethics Committee of National Taiwan University Hospital approved this study (approval number: 201410050RINB).