Co-infection of a single host by multiple virus genotypes or species is common in nature, facilitating studies of ecological interactions between viruses at the cellular level. When two or more viruses co-infect the same host cell, this can have profound consequences for the fitness (growth performance) of an individual virus. Co-infection may be advantageous to an individual virus due to increased pathogenesis, enhanced transmission, or the opportunity for genetic exchange (sex) which produces the raw material for natural selection. In contrast, co-infection may be disadvantageous to an individual virus because it increases the likelihood of competition for proteins and other resource products available within the cell. One intriguing cost of co-infection is the recent evidence that intra-cellular interactions between viruses can be antagonistic, where a virus genotype evolves to specialize in parasitizing other co-infecting viruses. Here I review laboratory experiments involving the RNA bacteriophage F6, which demonstrate the evolution of parasitism when viruses are propagated in environments where co-infection is common. Frequency-dependent selection is shown to govern the fitness of these parasitic (cheater) genotypes, because their benefit of cheating depends on the relative abundance of ordinary and cheater genotypes encountered within the host cell. I relate why the evolution of parasitic viruses may be relevant for observed limits to the absolute number of phages that can simultaneously infect a single cell. The evolution of parasitic interactions in viruses infecting animal and plant hosts is briefly discussed. I suggest directions for future research on the evolutionary ecology of virus co-infection, especially the need to study phage interactions under natural (non laboratory) conditions.