This study determined the effect of using mailing labels prepared by a letter-quality computer printer on survey response rate. D. A. Dillman's personalization approach to conducting mail surveys suggests that envelopes with addresses typed directly on them may produce a higher response rate than envelopes with addresses typed on self-adhesive labels. Two surveys were sent to the same 340 graduates of teacher preparation programs from one public university. In the first survey half of the graduates received up to three mailings in white No. 10 envelopes with addresses typed directly on them, while the other half received envelopes with addresses typed on self-adhesive mailing labels. The questionnaire contained four multiple choice items. Response rates were 80 percent and 78 percent respectively. In the second survey the same procedures were used, with half of each group from the first survey receiving the two types of envelopes. The questionnaire was longer and more complicated, and manila, instead of white No. 10, envelopes were used. The response rates were 63 percent for directly typed addresses and 65 percent for addresses on mailing labels. Results indicate letter-quality computer printers can be used to prepare address labels for surveys of an interested population without significantly reducing response rate. (Author/DWH)