A victim survey in southern Spain involved interviewing two independent samples by, respectively, face-to-face and telephone methods about their victimization experiences. The telephone interviews obtained a better response rate, while the face to face interviews showed higher crime rates. The survey found a strong inverse relationship between crime rates and response rates. The personal interviews were more difficult to carry out, and the refusal rate higher. However, those who conceded an interview had more to tell. A total of 76 percent experienced one or more crimes during the preceding 5 years, compared to 65 percent of those interviewed by telephone. The study also produced findings relevant to the issue of non-response. Persons who did not respond to the face-to-face interview were approached again, and asked to do the interview by phone instead. These initial non-responders reported fewer victimization events than the main sample. The article concludes that there is good reason to be suspicious of research attempts to compare findings that derive from different data collection techniques. Figure, references