The exclusive position of scientific expertise in pharmaceutical regulation is being increasingly challenged. Several authors suggest that lay knowledge could play a role in governing risks. We use the literature to develop ideal-typical regulatory arrangements with low and high lay stakeholder involvement: a technocratic and a democratic arrangement. We propose that a more technocratic arrangement will yield a better process and output performance while a more democratic arrangement will result in more stakeholder satisfaction. These propositions are explored through two case studies of pharmaceutical regulation in the Netherlands: in pandemic influenza and in HIV. Our study shows equivalent process and output performances but we found indications that the democratic approach results in more stakeholder satisfaction. We conclude that in pharmaceutical regulation, there is no a priori reason to limit involvement to experts: in situations of fundamental uncertainty, democratic monitoring of pharmaceutical risks can contribute to the system's robustness.