This 1983 study used a 'snowball' sampling technique to identify criminologist expert witnesses. This approach began with a list of the names and addresses of 11 'experts,' who were sent a questionnaire that provided space for the names and addresses of 'experts' known to the respondent. To qualify for the final sample, a respondent had to be a behavioral or social scientist who had testified as an expert witness on the basis of criminological knowledge. A total of 136 such persons were sent a questionnaire, with 90 usable questionnaires returned. A second sample consisted of persons with interests similar to the experts but who had no experience as expert witnesses. Ninety-nine usable questionnaire were received from this group. The same questionnaire items were used with both groups, but the order of the questions differed so as to move those with no experience as expert witnesses past contingency questions that dealt with experiences as expert witnesses. The questions were designed to measure the respondents' perceptions of the advocacy potential in testifying as an expert and the preparation required for such testimony. Questionnaire items were factor analyzed using the principal components method, and oblique rotation was used. Findings indicate that a pragmatic concern for pretrial preparation for testimony was greater among those with experience as expert witnesses. Also, respondents with expert witnessing experience were more advocacy oriented than those without such experience. Tabular data and 74 references are provided.