The outcome of the measurement of repeat victimization can influence how police agencies and others establish and assess crime control efforts. The results of measurement can also influence the theories and explanations resulting from empirical study. This study used data for 3 years from Baltimore, MD, Dallas, TX, and San Diego, CA, and included information on 20,000 to 40,000 burglaries in each city. The analysis compared rates of repeat victimization from data for different periods of time, using 36 data sets of 1 month each. Results revealed that a 1-year time window captured 42 percent more repeats than did a 6-month window and that a 3-year window captured 57 percent more repeats than did a 1-year window. Findings indicated that the rates of repeat victimization revealed by crime surveys were gross underestimates. The analysis concluded that the longer the time window, the better, and that methodological issues related to measuring crime rates could be important for policy, practice, and crime theory. Figure, table, and 17 references