Atomic scale computer simulations are used to investigate the intracascade evolution of the defect distributions of cascades in copper over macroscopic times. Starting with cascades generated using molecular dynamics, the diffusive transport and interactions of the defects are followed for hundreds of seconds in stochastic annealing simulations. The temperature and energy dependencies of annihilation, clustering and free defect production are determined for individual cascades. The simulation results illustrate the strong influence on intracascade evolution of the defect configuration existing in the primary damage state. Other significant factors affecting the evolution of the defect distribution are the large differences in mobility and stability of vacancy and interstitial defects and the rapid one-dimensional diffusion of small, glissile interstitial loops produced directly in cascades. The latter factor introduces a cascade energy dependence of defect evolution that is apparent only beyond the primary damage state, amplifying the need for performing many more cascade simulations at higher energies.