In this paper, we experimentally investigate the effect of public consumer ratings on market outcomes in credence goods markets. Contrary to search or experience goods, consumers can- not evaluate all dimensions of trade for credence goods, which may inhibit the information and reputation-building value of public rating systems. We implement a healthcare market frame in which physicians as experts have an informational advantage over patients with respect to the appropriate treatment. The rating system takes the form of a five-star rating system as is common on online rating websites. The value of this rating system is compared in two different expert market settings: First, one in which patients cannot rely on information from personal experience with the expert, reflecting markets in which consumer-expert interactions are often first-time and in- frequent (e.g. specialist visits). Second, one in which patients have personal experience with the expert, reflecting markets in which consumer-expert interactions are frequent and repeated (e.g. general practitioner visits). We find that the public rating system significantly improves market outcomes. Furthermore, a public rating system is a good substitute for personal experience in- formation in terms of market efficiency and consumer surplus. Combined, however, we find no complementarity between public ratings and personal experience information, mainly due to the already high market efficiency in the presence of either one.