The article explores epistemic and social conditions of the trustworthiness of scientific expertise. I claim that there are three kinds of conditions for the trustworthiness of scientific expertise. The first condition is epistemic and means that scientific knowledge enjoys high credibility. The second condition concerns the significance of scientific knowledge. It means that scientific generalizations are relevant for elucidating the particular cases that constitute the challenges for expert judgment. The third condition concerns the social processes involved in producing science-based recommendations. In this context trust is created by social robustness, expert legitimacy, and social participation.