We develop modeling and analysis approaches for quantifying the in vivo evolution of clonal blood cell populations. A stochastic birth-death-immigration (BDI) model is used to describe experiments that measure hematopoietic stem cell tags and T cell immunotypes. We investigate a number of important processes that affect the predictions of the BDI model, including stochasticity, subsampling, regulation, and heterogeneity in rates. Using variations of this underlying mathematical framework, we estimate the effective proliferative potential of progenitor cells, the heterogeneity of naive T cell output by the thymus, and the heterogeneity of their homeostatic proliferation rates.