INTRODUCTION: Fragility indices (FI) are a novel method to measure the robustness of results in clinical trials. The FI represents the number of participants whose status in a clinical trial would have to change from a non-event (not experiencing the primary endpoint) to an event (experiencing the primary endpoint) to turn a statistically significant result into a non-significant result. The aim of this study was to evaluate the robustness of trials for irritable bowel syndrome (IBS) including IBS-mixed, IBS-constipation, and IBS-diarrhea. METHODS: Trials published in high impact journals comparing the efficacy of therapies for IBS were identified from Medline. Trials had to be in adults, randomized, parallel-armed, with at least one statistically significant (P < 0.05) binary outcome, and a positive result (i.e. primary endpoint met). Only primary endpoints of therapeutic efficacy were considered. Only trials comparing Rifaximin, Eluxadoline, fecal microbiota transplant (FMT), probiotics, Ramosetron, Alosetron, physical activity, Cimetropium and Pinaverium to placebo met the inclusion criteria. FI and correlation coefficients were calculated and regression modeling was also used to identify predictors of a high FI. RESULTS: Thirteen trials meeting the inclusion criteria were analyzed. The impact factor of the journals in which they were published ranged from 3.5-79. The median FI was 6 (range 1-38). The median sample size in all trials was 427 (range: 44-1280). The trial publication year (P = 0.60), journal impact factor (P = 0.16), time point in weeks when the primary outcome was assessed (P = 0.45), duration of the study (P = 0.89), and number needed to treat [NNT] (P = 0.06) were not predictive of a high FI. However, a larger sample size was predictive of a higher FI (P = 0.01). The highest FI was seen in a Ramosetron trial (FI = 38) for IBS-D published in Gastroenterology. No correlation was seen between FI and journal impact factor (R = 0.19, P = 0.55), trial publication year (R = 0.09, P = 0.77), duration of trial (R = -0.45, P = 0.12), time point at which the primary endpoint was assessed (R = -0.12, P = 0.70), NNT (R = -0.34, P = 0.26), and sample size (R = 0.22, P = 0.47) CONCLUSION: Our results suggest that a median of six participants is needed to nullify efficacy outcomes in the included IBS therapy trials. A larger sample size was predictive of a high FI. Some trials used in the formation of evidence-based guidelines are fragile suggesting how easily statistical significance based on a threshold P-value may be overturned.