Baculoviruses have enormous potential for use as biopesticides to control insect pest populations without the adverse environmental effects posed by the widespread use of chemical pesticides. However, continuous baculovirus production is susceptible to DNA mutation and the subsequent production of defective interfering particles (DIPs). The amount of DIPs produced and their genome length distribution are of great interest not only for baculoviruses but for many other DNA and RNA viruses. In this study, we elucidate this aspect of virus replication using baculovirus as an example system and both experimental and modeling studies. The existing mathematical models for the virus replication process consider DIPs as a lumped quantity and do not consider the genome length distribution of the DIPs. In this study, a detailed population balance model for the cell-virus culture is presented, which predicts the genome length distribution of the DIP population along with their relative proportion. The model is simulated using the kinetic Monte Carlo algorithm, and the results agree well with the experimental results. Using this model, a practical strategy to maintain the DIP fraction to near to its maximum and minimum limits has been demonstrated.