Predicting microstructure evolution of metals under irradiation requires an in depth understanding of the production and properties of point defects and their clusters. Although, analytical descriptions based on the production bias model have recently been applied successfully to explain the behaviour of metals under different irradiation conditions, they still lack an important ingredient, namely that of the diffusion mechanism of the defects produced. In the present investigation, a hierarchical computer simulation model which takes into account spatial and temporal correlation in the production and diffusion of the defects is used to investigate the production of freely migrating defects in irradiated aluminium. The approach uses classical molecular dynamics with the embedded atom method potential to investigate the primary damage state from displacement cascades and a kinetic Monte Carlo simulation to follow defect kinetics over macroscopic time scales. In the present paper, Re have used this approach to investigate the primary damage state and the fraction of freely migrating defects in pure Al as a function of irradiation particle energy and crystal temperature. The results ale compared to other FCC metals and to the experiments when available. (C) 1999 Elsevier Science B.V, All rights reserved.