The findings show that respondent's predictions of future victimization were not usually related to the frequency or nature of prior victimizations, and the predictions were not typically accurate. Generally, respondents predicted that they would be crime victims within the next year. Reasons for such pessimistic predictions are unclear in survey responses. Based on this research, the only conclusion offered is that respondent's predictions about future victimization were not based on a rational calculation based on past experience or an assessment of mathematical probabilities related to measured crime rates in areas where they lived and worked. The authors suggest that future surveys inquire about the reasons for a respondent's prediction of future victimization. Data were obtained from the three waves of Trinidad's Community Living and Integration Survey. The first full wave was conducted in September 1999 (n=728), the second in September 2000 (n=636), and the third in September 2001 (n=716). A total of 330 respondents were surveyed in all 3 waves; they were the sources of data used in the current study. Questions that were the focus of this study addressed predictions of experiencing the following events in the next year: being approached by a beggar, being cheated out of money, being burglarized, being raped or sexually assaulted, being murdered, being attacked with a weapon, being robbed or mugged, being vandalized, being kidnapped, being victimized in the workplace, and being a victim of any crime in the near future. 3 tables, 1 figure, and 27 references