Study Objectives It is known that antidepressant drugs can induce sleep disorders in patients, but little data exist about high or low-risk molecules. The aim was to study the frequency of antidepressant drugs-induced sleep disorders (DISD) by molecule. Methods 77,391 patient comments for 32 antidepressant drugs were collected from drug review websites and screened for DISD. Association between drugs and nightmare disorder, restless legs syndrome, sleep paralysis, sleep terrors, sleep-related hallucinations or sleep walking was expressed as relative proportion [proportional reporting ratio (PRR)]. A detailed analysis of the dreams content was also carried out. Results Amitriptyline, doxepin, fluvoxamine, mirtazapine, nortriptyline, trazodone, venlafaxine and vilazodone were associated with a greater frequency of DISD compared to other antidepressants. Vilazodone heavily increased the probability of developing 5 of the 6 studied DISD (PRR 3.3 to 19.3) and mirtazapine increased the probability for developing 4 DISD (PRR 2.4 to 6.4). Bupropion and citalopram were associated with lower probabilities for 5 DISD (PRR 0.2 to 0.7). Sentiment analysis showed that patients described disturbing dreams for vilazodone or mirtazapine and strange but less negative dreams for bupropion, citalopram or duloxetine. Conclusions Relative frequencies of sleep disorders were obtained for a vast panel of antidepressant drugs through an original analysis of user’s drug reviews on drug rating websites. Our results could guide clinicians in appropriate choice of antidepressant drug for high DISD-risk patients in need of such treatment. These results may however be cautiously taken, considering the uncertain reliability and generalisability of web-based data.