Factors affecting 750 Wyoming public school teachers' response rates to a mail survey examined in this study were: personalization of the cover letter and of the envelope; grade level taught (elementary, junior, or senior high); sex of teacher; and regional characteristics (rural area, small town, or large town). Two experimental treatments were used, varying the personalization used in mailing the questionnaire. The six-page instrument concerned the use of achievement tests in the classroom. Two follow-ups were sent. The overall response rate was an acceptable 80 per cent. A log linear model was used to analyze the effects of five variables on survey returns: personalized salutation, original signature, grade level taught, sex, and population of the school's location. The response rate was significantly higher from rural areas and small towns than from larger towns. The personalization aspects did not significantly affect response rates, but an interactive effect was suggested. There were no significant effects of sex or of grade level taught. However, there were differential responses to the first, second, and third mailings by geographic locale, sex, and grade level taught. General conclusions regarding respondent characteristics would be potentially dangerous, without considering the number of follow-ups as a factor. (GDC)