This thesis is an investigation of the metaphysical implications of possible worlds semantics. It has three aims: first, to present a case for taking possible worlds seriously; second, to present several plausible candidates for the role ascribed to possible worlds; third, to chart out some of the main lines of debate. I focus my discussion specifically on the full-blooded worlds of David Lewis, the world-stories of Robert Adams and the world-natures of Robert Stalnaker and Peter Forrest. After laying out each theory, I consider several main objections to each. I argue that the world-story theory and the world-nature theory stand up well to their critics. Ultimately, I find that the world-nature theory is superior as it is more able to cope with the problems that its critics highlight.