Internet shares some characteristics of survey making with traditional media, especially postal mail. However, there are considerable differences that justify a different focus on administration and make existing knowledge of the traditional media not directly applicable to Internet. Research is therefore necessary to discover how Web-based surveys operate under different conditioning factors, so that general behavioral patterns can be established in order to improve the administration and results of such surveys. This study thus centers on two of the parameters that can influence responses to Web-based surveys, which are personalization and the frequency of reminder mailings distributed among the sample population. The results obtained show a positive influence of personalized e-mail messages on response rate and the need to use a lower frequency for studies aiming at increasing the response rate in the shortest possible time; and longer frequency (and personalized) when the aim is to have the respondents complete the full questionnaire.