It is often thought that the existence of other worlds cannot be scientifically verified and therefore should be treated as philosophical speculation. In this article, I describe several methods for determining if other worlds exist, even without interacting with them. These methods are based on the following premise: if there are many worlds, then the statistical properties of a natural process are biased when measured by an observer whose existence was influenced by the process. The bias is always in the same direction, making the process appear more beneficial for the existence of the observer than it actually is. I suggest several potential ways of measuring the bias, show through a simple model of population dynamics how the bias is generated, and briefly consider whether our current drop in population growth is evidence of many worlds.