Functions that are to be minimized by derivative-free algorithms are frequently noisy. Whether this noise is stochastic or deterministic, it can greatly influence the progression of a derivative-free routine. In this talk, we highlight applications of derivative-free optimization to noisy problems as well as provide an overview of previous techniques for solving these problems. We propose handling the noise, not by repeated sampling of the function, but rather by constructing models around points of interest. Lastly, we introduce solver-independent termination criteria for algorithms running on noisy problems and propose tools to measure the quality of the termination point.