For the complex problems faced by policy-makers today, it is rare that evidence drawn from a single discipline is sufficient for deciding upon a course of action. Instead, it is usually necessary, even optimal, to draw on evidence from multiple disciplines, thus producing multiple perspectives on an issue. Although in the ideal situation, the evidence will all point to the same ultimate conclusion, in practice, such neat convergence is rare. More common are truly heterogeneous sets of evidence, drawn not only from different scientific disciplines, but also indicative of different claims about the subject. For example, when faced with a potentially toxic substance, one will turn to studies from animal toxicology, human epidemiology, and biochemistry, and often some studies will indicate a positive link between a substance and some adverse effect (though the particular effect indicated may not be consistent across studies), and some will not. In this symposium, experts from different relevant fields will describe and critique approaches to grapple with such complexity. The symposium will provide a constructive answer to the question, "How should we weigh divergent and internally inconsistent sets of evidence?"