Every evolutionary biologist will surely acknowledge that convergent evolution, the independent evolution of similar features in different evolutionary lineages, is an important phenomenon of the organic evolution. However, the concept is complex and can have several related but conceptually distinct meanings in the literature, including parallel evolution (independent mutations in orthologous genes) and homoplasy (any convergent traits, including reversion). Some authors, for example, use the term “parallel evolution” differently from “convergent evolution”, and they reserve the latter term for more “unlikely” (or “more independent”) examples of phenotypic similarity across lineages, those not predisposed by genomic similarity. Semantic arguments in science are often fruitful, but can also prevent efficient scientific exchanges in the field. Hence, the goal of this article was (1) to define convergent evolution in a better way by applying a multilevel biological-level approach and (2) to propose a road map to help researchers navigate their routes in studying this phenomenon.