Within 100 years after the Spanish conquest, the Amerindian population of the New World was reduced to a small fraction of its former size. While the fact of this depopulation is not seriously disputed, virtually everything else about it is (e.g., the size of the pre-Colombian population, the scale of its decline, and the relative importance of various causes). These questions are examined using a system dynamics computer simulation methodology applied to a case study in the sixteenth-century Basin of Mexico. This method simulates the demographic response of a population to the web of causal factors that determine population size. These factors constitute a structured cultural ecology of the Amerindian population, and include demographic, epidemiological, cultural/social, and productive aspects of the Basin's population system. The results of these simulations indicate that: (1) very large depopulation was possible given reasonable assumptions as to cause; (2) the overall scale of depopulation was profound-nearly 90 percent over the course of 100 years-from more than 1.5 million in 1519 to less than 200,000 by 1610; (3) the temporal pattern of depopulation formed an irregular, step-like pattern and most of the population loss occurred in the first fifty years of Spanish occupance; (4) this decline was primarily due to a series of "virgin soil" epidemic crises, although famine was also important; and (5) there is no need to presume an Amerindian population genetically less able to resist disease, nor is it necessary to presume Spanish cruelty to explain this holocaust. This study provides independent support for previous studies that posit a catastrophic population collapse in the sixteenth century and refutes arguments for either a more extreme collapse or a lesser decline. By methodologically connecting the causes of the population decline with the results (the population statistics), it illuminates debate over the relative importance of different proposed causes.