Summary Background Although dietary emulsifiers are implicated in the pathogenesis of Crohn's disease, their effect has not been studied in humans. Aim To determine the effects of high‐ and low‐emulsifier diets (HED, LED) on intestinal barrier function in healthy subjects in unstressed and acutely stressed states. Methods We conducted a single‐blinded, cross‐over, controlled feeding trial in 22 healthy adults. After recording 7 days of their habitual diet, we randomised participants to HED or LED with ≥3‐week washout between diets. On dietary completion, acute stress was induced via intravenous corticotrophin‐releasing hormone. We assessed dietary adherence, effects on 2‐h urinary lactulose: rhamnose ratio (LRR), serum concentrations of lipopolysaccharide‐binding protein, soluble‐CD14 and markers of epithelial injury and inflammation. Results Dietary adherence was excellent. In an unstressed state, median (interquartile range) LRR during HED was 0.030 (0.018–0.042); on LED, this was 0.042 (0.029–0.078; p = 0.04). LPB concentrations were lower on HED than LED ( p = 0.026), but no differences were observed for epithelial injury or inflammation. Under acute stress, LRR increased by 89% (−1% to 486%) on HED ( p = 0.004), differing ( p = 0.001) from 39% (1%–90%) decrease on LED ( p = 0.009). Soluble‐CD14 also increased ( p < 0.001). The LED had a prolonged carry‐over effect on suppressing HED‐induced changes during stress. Similar changes in LRR and soluble‐CD14 were observed when HED was used as the first diet (both p < 0.01). Conclusion High intake of emulsifiers improved barrier function in the unstressed state, but increased intestinal permeability to stress, without evidence of inflammation. A LED was protective of the stress effect.