Randomized optimization heuristics (ROHs) are general-purpose solvers that are frequently applied for discrete or continuous optimization when other classes of solvers fail or are not applicable. Typical examples of ROHs include genetic algorithms, evolution strategies, and estimation-of-distribution algorithms (EDAs). Since ROHs are not designed with a proof or analysis in mind, their theoretical analysis is a particularly challenging endeavor. Yet, their practical success, extremely wide applicability, and growing relevance in many domains in industry and academic research push for progress in their understanding. This Dagstuhl Seminar is part of a very successful Dagstuhl Seminar series with the goal of driving this research area forward.