Starch microparticles have been shown to be effective as a particulate adjuvant carrier in oral vaccination. In mice, formulations with bacterial antigens have elicited both systemic and mucosal immune responses providing protection upon challenge with live bacteria. A vaccine formulation with formaldehyde-treated diphtheria toxin cross-reacting material, CRM197, optimised in mice, was tested in healthy volunteers in a booster design. Specific antibodies as well as toxin-neutralising antibodies in a Vero cell analysis indicated that the vaccine was not effective in man. It is obvious that the longer transit time in the human GI tract and possible unfavourable distribution of Peyer's patches and M-cells necessary for the uptake of the starch particles require a more stable formulation. It is proposed that enteric coating of the particles or particles in a gastro-resistant capsule could be a more efficacious vaccine formulation.