Growing empirical literature has sought to explain the forces behind the significant expansion of the United States prison population over the past 30 years. However, the studies to date have suffered from important methodological, conceptual, and definitional problems that have curtailed their ability to identify causal effects. Many studies are now focusing on this significant issue shedding light on the driving forces behind the growth. In order to generate more accurate results, several steps can be taken, such as (1) closer attention be given to issues of endogeneity with respect to crime rates, unemployment, sentencing guidelines, and other factors, (2) more care is required in confronting sensitivity to functional form, (3) the particular choice of variables demand more consideration, prison admissions and releases need to replace total prison population as the dependent variable, and (4) studies should seek out alternate sources of data, either looking to different jurisdictions or different sets of data. The current models may not have answered the question of population growth fully, but they have begun an important process. With more powerful techniques, better appreciation of particular structural problems, clearer understandings of how the data look, and careful selection of relevant explanatory variables, it is possible to derive increasingly reliable estimates of the forces shaping prison populations today. This article has two goals: provide a systematic analysis of what is known about the forces that have driven prison population growth and to set forth a framework for producing more reliable and accurate results in the future. Figures and tables