The benefits of police data are sometimes forgotten by victimologists; it is the total population of recorded offenses. The biggest limitation of police data is the question of crimes which are not reported. In addition, because police data rely on administrative process and legislative factors, one cannot make accurate comparisons between times or areas. The major benefit of survey data is that they include offenses which are not reported to police and the survey can be repeated to furnish comparisons over time or between geographic areas. Survey data also can provide more details about victims, circumstances and consequences of crimes. The limitations of surveys arise from the fact that they are a sample and as such will be subject to sampling errors. Any results are also subject to confidence limits and problems with the recall and memory of interviewees; they generally exclude children and businesses; and they usually limit the type of offenses under review. The article proposes that the two sources of information can complement each other and uses results of a survey of violence against women in Australia to illustrate the point. References