BACKGROUND: Environmental enteropathy (EE) is an asymptomatic abnormality of small bowel structure and function, which may underlie vaccine inefficacy in the developing world. HIV infection co-exists in many of these populations. There is currently no effective treatment. We conducted a secondary analysis of a randomised controlled trial of high dose multiple micronutrient (MM) supplementation on small bowel architecture in EE in participants with or without HIV infection. METHODS: In a double-blind parallel-group trial of the effect of MM on innate immune responses to oral vaccines, consenting Zambian adults were randomised to receive 6 weeks of 24 micronutrients as a daily capsule or placebo. HIV status was established after randomisation. Proximal jejunal biopsies were obtained after the supplementation period. Villous height, crypt depth, villous width, villous perimeter per 100 μm muscularis mucosa (a measure of epithelial surface area), and villous cross sectional area per 100 μm muscularis mucosa (a measure of villous compartment volume) were measured in orientated biopsy sections using semi-automated image analysis. Analysis was by intention to treat. RESULTS: 18 patients received MM and 20 placebo. 6/18 MM and 9/20 placebo patients had HIV. In HIV negative patients given MM compared to placebo, mean villous height was 24.0% greater (293.3 v. 236.6 μm; 95% CI of difference 17.7-95.9 μm; P = 0.006), mean villous area was 27.6% greater (27623 v. 21650 μm2/100 μm; 95% CI of difference 818-11130 μm2/100 μm; P = 0.03), and median villous perimeter was 29.7% greater (355.0 v. 273.7 μm/100 μm; 95% CI of difference 16.3-146.2 μm/100 μm; P = 0.003). There was no significant effect on crypt depth or villous width. No effect was observed in HIV positive patients. There were no adverse events attributable to MM. CONCLUSIONS: MM improved small bowel villous height and absorptive area, but not crypt depth, in adults with EE without HIV. Nutritional intervention may therefore selectively influence villous compartment remodelling. In this small study, there was a clear difference in response depending on HIV status, suggesting that EE with superimposed HIV enteropathy may be a distinct pathophysiological condition.