INTRODUCTION: Evidence about specific carbohydrate diet (SCD) for inflammatory bowel disease (IBD) is limited. We conducted 54 single-subject, double-crossover N-of-1 trials comparing SCD with a modified SCD (MSCD) and comparing each with the participant's baseline, usual diet (UD). METHODS: Across 19 sites, we recruited patients aged 7-18 years with IBD and active inflammation. Following a 2-week baseline (UD), patients were randomized to 1 of 2 sequences of 4 alternating 8-week SCD and MSCD periods. Outcomes included fecal calprotectin and patient-reported symptoms. We report posterior probabilities from Bayesian models comparing diets. RESULTS: Twenty-one (39%) participants completed the trial, 9 (17%) completed a single crossover, and 24 (44%) withdrew. Withdrawal or early completion occurred commonly (lack of response [n = 11], adverse events [n = 11], and not desiring to continue [n = 6]). SCD and MSCD performed similarly for most individuals. On average, there was <1% probability of a clinically meaningful difference in IBD symptoms between SCD and MSCD. The average treatment difference was -0.3 (95% credible interval -1.2, 0.75). There was no significant difference in the ratio of fecal calprotectin geometric means comparing SCD and MSCD (0.77, 95% credible interval 0.51, 1.10). Some individuals had improvement in symptoms and fecal calprotectin compared with their UD, whereas others did not. DISCUSSION: SCD and MSCD did not consistently improve symptoms or inflammation, although some individuals may have benefited. However, there are inherent difficulties in examining dietary changes that complicate study design and ultimately conclusions regarding effectiveness.