Study goals were to examine statistically the contention that response patterns to mailed surveys are stable and predictable and to provide a statistical procedure to forecast anticipated response rates based on certain statistical assumptions. Data analysis was based on the premise that if patterns of response are relatively stable and predictable, a statistical distribution can be found to fit the observed patterns of response. Analysis of the 38 surveys indicates that the gamma distribution can be used to construct a profile of response patterns, as determined by the statistical test of Kolmogorov-Smirnov. Based on this finding, researchers could forecast the patterns of response to mail surveys from the mean and variance of the distribution of daily questionnaire returns. Data required for forecasting include the mean value of questionnaire returns per day, the variance of questionnaire returns, and the overall response rate. Results suggest that examination of the daily probabilities of questionnaire returns may aid the development of an effective call-back procedure to nonrespondents and can help minimize the waiting time for survey completion. Figures, tables, notes, and four references are provided. (Author abstract modified)