Naive CD4 and CD8 T cells are part of the foundation of adaptive immune responses, but multiple aspects of their behaviour remain elusive. Newly generated T cells continue to develop after they leave the thymus and their dynamics and 'rules of entry' into the mature naive population are challenging to define. The extents to which naive T cells' capacities to survive or self-renew change as they age are also unclear. Further, much of what we know about their behaviour derives from studies in adults, both mouse and human. We know much less about naive T cell dynamics early in life, during which the thymus is highly active and peripheral T cell populations are rapidly established. For example, it has been suggested that neonatal mice are lymphopenic; if so, does this environment impact the behaviour of the earliest thymic emigrants, for example through altered rates of division and loss? In this study we integrate data from multiple experimental systems to construct models of naive CD4 and CD8 T cell population dynamics across the entire mouse lifespan. We infer that both subsets progressively increase their capacity to persist through survival mechanisms rather than through self-renewal, and find that this very simple model of adaptation describes the population dynamics of naive CD4 T cells from birth into old age. In addition, we find that newly generated naive CD8 T cells are lost at an elevated rate for the first 3-4 weeks of life, which may derive from transiently increased recruitment into conventional and virtual memory populations. We find no evidence for elevated rates of division of naive CD4 or CD8 T cells early in life and indeed estimate that these cells divide extremely rarely. Markers of proliferation within peripheral naive T cells are instead inherited from division during thymic development. We also find no evidence for feedback regulation of rates of division or loss of naive T cells at any age in healthy mice, challenging the dogma that their numbers are homeostatically regulated. Our analyses show how confronting an array of mechanistic mathematical models with diverse datasets can move us closer to a complete, and remarkably simple, picture of naive CD4 and CD8 T cell dynamics in mice.