This review focuses on the effects of multiple contacts with respondents on studies of response rates to mail surveys. After screening articles on research criteria, such as the presence of original research and sufficient reporting detail, 28 studies remained of the original 208 identified. The difference in response rate between treatment and control groups (or comparison groups) was calculated for each point in the process, such as initial and followup contacts. In 16 cases, the difference in response rate between treatment and control (or between treatment groups) was increased when a second contact was instituted, In 27 cases, however, it was diminished, and in 7 cases it reversed. Findings tend to be inconclusive, but it can be seen that followups have the potential for confounding the results of experiments aimed at determining the most effective procedures for facilitating response rates in mail surveys. The most consistent finding is that one followup is more likely to decrease the effect of the manipulation than to enhance or reverse it. However, the magnitude of group differences must be considered along with the effect of the additional contact. With the limited number of cases available, it is not possible to draw hard and fast conclusions regarding many of the independent variables studied or other study characteristics. The 28 studies cited in the review are listed. (Contains 22 references.) (SLD)