The interplay between atomic and nuclear interactions in collisions between light and heavy ions is studied. The general theoretical description is outlined and analysed in a number of different limits (non-relativistic and relativistic electrons, semiclassical approximation, DWBA, fully quantal description) that have been used in practical applications. The two most important physical mechanisms for generating atomicnuclear interference, i.e. energy conservation and the introduction of additional phase shifts by nuclear reactions, are extracted and their universality in all scattering systems is stressed. The need for choosing different sets of basis states in light and heavy, symmetric and asymmetric systems, in order to achieve an economical theoretical description, is discussed in some detail. The resulting typical coupling matrix elements are analysed for their relative importance in atomic and nuclear excitations. The description of nuclear influence on atomic excitations in terms of a classical time delay caused by nuclear reactions is reviewed and its relationship to the underlying quantal character of the nuclear reaction is extracted.An experimental chapter reviews the existing data on nuclear-atomic interference. A large section covers proton-induced reactions (through isobaric analogue resonances or compound-nucleus formation) and their effects on bremsstrahlung spectra, K-hole probabilities, x-ray emission and electron capture. Inverse processes, where excitations within the electron shell modify the nuclear scattering, are also described. That structures in the atomic spectra can be consistently used to extract nuclear reaction times or average nuclear resonance widths is demonstrated for compound-nucleus formation in light-ion scattering and later applied to heavy-ion reactions. The experimental use of K-hole, 8-electron and positron production in heavy-ion reactions as an atomic clock for nuclear delay times is discussed; it is shown to give consistent results even in a region where direct determination of these times from measurement of nuclear reaction products is so far impossible.A prominent place in this review is reserved for spontaneous positron emission in supercritical heavy-ion collisions (Z,,, B 173); line structures predicted in the positron energy spectra may provide the most sensitive mechanism to isolate even very small fractions of the nuclear cross section proceeding through strongly delayed resonances or reaction channels. Attempts to use this idea to interpret the recently discovered