Abstract:The search for weak-interaction induced atomic parity non-conservation, initiated in the 70s, challenges both theory and experiment. Since the weak interaction is very short-range, the atomic effects increase rapidly with nuclear charge, as ¢ ¡. The focus has thus been on heavy atoms, where relativistic effects are essential, and nuclear size must be taken into account. The generalization of atomic many-body methods to relativistic systems involved both computational and formal difficulties, incorporating methods developed in quantum electrodynamics. Twenty years ago, the ability was emerging to treat atomic pair correlation in a relativistic framework. The application to many-electron systems opened for comparison with many atomic properties, such as isotope shifts, hyperfine structure and hyperfine anomalies, which reflect nuclear properties. In addition, the search for simultaneous violation of both parity and time reversal symmetry involves different types of effects, including nuclear "Schiff moments". Comparison between computed and experimental results for highly charged hydrogen-like systems provides a test of the theoretical and numerical treatment of the electron-nucleus interaction and of the description of nuclear distributions.