In this Opinion piece, I argue that the dynamics of viruses and the cellular immune response depend on the body size of the host. I use allometric scaling theory to interpret observed quantitative differences in the infection dynamics of lymphocytic choriomeningitis virus (LCMV) in mice (Mus musculus), simian immunodeficiency virus (SIV) in rhesus macaques (Macaca mulatta) and human immunodeficiency virus (HIV) in humans.It is well-known that the average metabolic rate of cells is typically lower in larger species (Kleiber, 1932). Allometric theory predicts that the total metabolic rate of an organism scales approximately as M^(3/4), with M being the body mass (West et al., 1997). Assuming invariant size and volume of cells between species, the metabolic rate of a single infected cell then scales as M^(-1/4). Other properties such as the lifespan of an animal or the number of certain cell types have also been found to depend on body size (Peters, 1983; Schmidt-Nielsen, 1984; Calder, 1996; Savage et al., 2007). In host-pathogen systems, the following three aspects could underlie allometric scaling laws (Wiegel and Perelson, 2004; Cable et al., 2007; Banerjee and Moses, 2010). First, the metabolic rate will affect the rate at which cells synthesize DNA and proteins and could therefore influence the replication rate of viruses. Second, quantitative processes of the cellular immune response could be affected by allometric scaling. Third, the interaction between viral replication rates and the dynamics of immune responses could result in differences in the time to disease progression between host species.Insights of the within-host dynamics of HIV is often based on experimental studies in macaques. Despite being a non-natural host, macaques can be infected with SIV, the simian counterpart of HIV. The viral turnover within a host has been found to be more rapid in macaques than in humans, with estimated half- lives of virus-producing cells of 0.5 days and 0.7 days, respectively (Markowitz et al., 2003; Brandin et al., 2006). In this light, it is interesting to compare the dynamics at which immune escape variants evade recognition from CD8+ cytotoxic T lymphocyte (CTL) responses in the two host species. Asquith and McLean (2007) showed that escape rates in macaques are about twice as fast as in humans, and suggested more efficient CTL-mediated killing of infected cells in macaques. An alternative hypothesis would be that the rate of immune escape depends on the overall viral turnover (Althaus and De Boer, 2012). With a 10-fold difference in body mass between humans (∼70 kg) and macaques (∼7 kg), allometric scaling predicts a 1.8-fold difference in the metabolic rate of a single cell between the two species. This factor is close to the reported 2-fold difference in escape rates and the 1.4-fold difference in the viral turnover between humans and macaques. This suggests that the metabolic rate could indeed directly affect the rate of virus replication and indirectly influence the rate of immune escape. However, it is worth noting that rates of CTL escape show substantial variation within host species and are generally faster during the acute phase of infection compared to the chronic phase (O’Connor et al., 2002; Asquith et al., 2006; Althaus and De Boer, 2008).Previous studies have investigated how quantitative processes of the cellular immune response could be affected by allometric scaling. In the case of T cell responses, Wiegel and Perelson (2004) derived some general principles on how the number of naive T cells scales with body size. Lymphocyte trafficking, i.e., the circulation of T cells through blood, tissues and the lymphatic system in order to …