The interactive effects of dietary protein level, zinc oxide (ZnO) supplementation and infection with an enterotoxigenic strain of Escherichia coli (ETEC) on performance responses and gastrointestinal tract characteristics were examined. Ninety-six individually housed, 21-day-old pigs (1 : 1 gender ratio) with initial bodyweight (BW) of 7.2 ± 0.69 kg, were used in a split plot experiment, with the whole plot being challenge or no challenge with ETEC and the dietary treatments used as subplots and arranged in a completely randomised 2 × 2 factorial design, with the factors being (i) two dietary protein levels (251 versus 192 g/kg crude protein) and (ii) addition or no addition of 2.5 g/kg ZnO. No antibiotic was added to the diet. The ETEC infection decreased average daily gain (P < 0.001) and increased feed conversion ratio (P < 0.01). Protein level had no effect on performance of pigs while ZnO supplementation increased (P < 0.001) average daily gain and average daily feed intake and hence decreased feed conversion ratio (P < 0.001). There were no 2- or 3-way interactions for growth performance indices (P > 0.05). Feeding a lower protein diet did not influence (P > 0.05) faecal volatile fatty acid concentrations. In non-infected pigs, feeding a lower protein diet caused a lower pH in the jejunum and ileum compared with pigs fed a higher protein diet (P < 0.05 and P < 0.01, respectively). However, feeding ZnO-supplemented diets increased (P < 0.05) the pH in the stomach and caecum compared with feeding diets without ZnO supplementation. Protein level did not alter (P > 0.05) empty BW but dietary supplementation with ZnO increased empty BW (P < 0.05). Neither protein level nor ZnO supplementation modified small intestinal morphology, although a tendency for an interaction (P < 0.1) was detected for jejunal villous height between protein level and ZnO supplementation. The results indicate that feeding ZnO-supplemented diets improved pig performance, and feeding a lower protein diet without ZnO supplementation did not compromise performance nor modify measures of gastrointestinal tract structure and function compared with pigs fed a diet higher in protein after weaning.