Advances in generating controlled few-cycle laser pulses and novel ultrashort XUV/Xray sources, from free electron laser (FEL)-based to attosecond high harmonic generation (HHG)-based, have opened completely new avenues for imaging electronic and nuclear dynamics in molecules, with exciting applications in physics, chemistry and biology. Processes such as ionization and dissociation of simple diatomic molecules can now be monitored in real time, but the access to few-femtosecond or attosecond time scales in the XUV/X-ray domain may also allow one to uncover and control the dynamics of elementary chemical processes such as, e.g., ultrafast charge migration, proton transfer, isomerization or multiple ionization, and to address new key questions about the role of attosecond coherent electron dynamics in chemical reactivity. The success of current experimental efforts in explaining these phenomena, present in many biological processes, is seriously limited due to the difficulty in their interpretation. In this respect, the implementation by the applicant’s group of nearly exact theoretical methods in supercomputers has made it possible to guide experimental research on simple systems. Such theoretical methods lie outside the traditional quantum chemistry realm since, e.g., they must accurately reproduce the time evolution of the coupled electronic and nuclear motions in the electronic and dissociative continua, including electron correlation and non-adiabatic effects. The necessary extension to systems of chemical interest, the current bottleneck in this field, requires extensive and novel theoretical developments along a similar direction. The aim of this project is to study the electronic and coupled electronic-nuclear dynamics in complex molecules at the attosecond or few-femtosecond time-scales, developing concepts and accurate theoretical tools to interpret the new generation of time-resolved experiments and to achieve ultrafast electronic control in chemistry.