I wrote the computer simulation model PARPOP to investigate the population dynamic of the Puerto Rican Parrot (Amazona vittata), a species with < 50 individuals in the wild. The matrix model is age-class based and incorporates the effects of hurricanes on mortality and recruitment. Under the present intensive management, the probability of extinction for the population over a one hundred year period is 4.7%, with an average population of 206 birds after 100 years. Using actual fledging data for 1985-2000, the model adequately predicted the overall trends in the population but over-predicted population size through the 1990's. Population estimates, however, are presented without error values, making an evaluation of the accuracy of the model difficult. The modeled population was most sensitive to changes in mortality of older breeding individuals followed, closely by the proportion of females that attempt to nest. The model highlights the importance of maintaining, and increasing if possible, the intensive management of this species for its survival. Moreover, the modeling highlights the necessity for greater quantitative rigor in population estimation and in the estimates of survival for different age classes. To accomplish this, it is suggested that radio tagging be used since the techniques for capturing and radio-tagging parrots are well researched, have been successfully applied to the Puerto Rican Parrot and other species of Amazona, and the value of the data out-weigh the minimal risk associated with the technique.