Introduction.Management of steroid-refractory ulcerative colitis has predominantly involved treatment with systemic cyclosporine A (CyA) and infliximab.Aim.The purpose of this study was to assess the effect of using a colon-targeted delivery system CyA formulation on the composition and functionality of the gut microbiota.Methodology.Ex vivofaecal fermentations from six healthy control subjects were treated with coated minispheres (SmPill) with (+) or without (−) CyA and compared with a non-treated control in a model colon system. In addition, thein vivoeffect of the SmPill+CyA formulation was investigated by analysing the gut microbiota in faecal samples collected before the administration of SmPill+CyA and after 7 consecutive days of administration from eight healthy subjects who participated in a pilot study.Results.Analysis of faecal samples by 16S rRNA gene sequencing indicated little variation in the diversity or relative abundance of the microbiota composition before or after treatment with SmPill minispheres with or without CyAex vivoor with CyAin vivo. Short-chain fatty acid profiles were evaluated using gas chromatography, showing an increase in the concentration of n-butyrate (P=0.02) and acetate (P=0.32) in the faecal fermented samples incubated in the presence of SmPill minispheres with or without CyA. This indicated that increased acetate and butyrate production was attributed to a component of the coated minispheres rather than an effect of CyA on the microbiota. Butyrate and acetate levels also increased significantly (P=0.05 for both) in the faecal samples of healthy individuals following 7 days’ treatment with SmPill+CyA in the pilot study.Conclusion.SmPill minispheres with or without CyA at the clinically relevant doses tested here have negligible direct effects on the gut microbiota composition. Butyrate and acetate production increased, however, in the presence of the beads in anex vivomodel system as well asin vivoin healthy subjects. Importantly, this study also demonstrates the relevance and value of usingex vivocolon models to predict thein vivoimpact of colon-targeted drugs directly on the gut microbiota.