This study reports on a field experiment in which the effects of manipulating the contents of the first reminder in an industrial survey were investigated. Every second non-responding company received a replacement questionnaire with the first mailed reminder, while the second reminder was identical in all cases. The study compared the impact on response rates, costs, timeliness and quality. The response rate between the first and second reminders was significantly higher in the replacement group, but the second reminder evens out most of this difference. In addition, logistic-regression models revealed significant differences in the reaction from different industries and regions. For the replacement group, the cost/response rate was 20% higher and the marginal cost per extra response was four times that of the responses to the first mailing. However, the questionnaires were returned faster from the replacement group up until the second reminder was launched and the quality was higher.