This chapter discusses the case for and against semiclassical radiation theory. Although semiclassical theories have had considerable success in accounting for many observed effects, including photoelectric correlations, pulse propagation through a resonant medium, photon echoes and optical free induction, they fail completely in other cases, and no evidence exists that should cause one to think of giving up quantum electrodynamics (Q.E.D.) in favor of a semiclassical theory. Despite the internal difficulties of Q.E.D. that remain to be resolved, all existing experimental evidence points rather to the validity of Q.E.D., whose predictions have never actually failed, and to the inadequacy of the semiclassical approach in certain situations. Of the various semiclassical approaches, neoclassical theory is the most internally self-consistent and divergence-free theory. Moreover, its predictions frequently differ from those of Q.E.D. and can, therefore, be put to the test, whereas much of the work on zero-point field theory has been aimed at reproducing the results derived by Q.E.D. The chapter, therefore, emphasizes the evidence relating to neoclassical theory although some of this evidence relates equally to all semiclassical theories.