Adaptive radiation refers to diversification from an ancestral species that produces descendants adapted to use a great variety of distinct ecological niches. In this review, I examine two aspects of adaptive radiation: first, that it results from ecological opportunity and, second, that it is deterministic in terms of its outcome and evolutionary trajectory. Ecological opportunity is usually a prerequisite for adaptive radiation, although in some cases, radiation can occur in the absence of preexisting opportunity. Nonetheless, many clades fail to radiate although seemingly in the presence of ecological opportunity; until methods are developed to identify and quantify ecological opportunity, the concept will have little predictive utility in understanding a priori when a clade might be expected to radiate. Although predicted by theory, replicated adaptive radiations occur only rarely, usually in closely related and poorly dispersing taxa found in the same region on islands or in lakes. Contingencies of a variety of types may usually preclude close similarity in the outcome of evolutionary diversification in other situations. Whether radiations usually unfold in the same general sequence is unclear because of the unreliability of methods requiring phylogenetic reconstruction of ancestral events. The synthesis of ecological, phylogenetic, experimental, and genomic advances promises to make the coming years a golden age for the study of adaptive radiation; natural history data, however, will always be crucial to understanding the forces shaping adaptation and evolutionary diversification.