Providing a respondent with a choice of modes for submitting a questionnaire would appear to improve response rates and thus reduce nonresponse. A survey of registered recreational boat owners was conducted using different modes and mode choice in 2009. The purpose of this survey was to document the yearly use and gasoline consumption of Oregon registered boats. We randomly assigned three mode groups to investigate whether providing a mode choice had an impact on reducing nonresponse. One group received printed versions of the questionnaire and was asked to return the completed form by mail. This group was not given a choice to complete the questionnaire by any mode other than mail. A letter was mailed to the second group providing them a link to a Web site with the questionnaire. No printed version of the questionnaire was sent with the first mailing. Nonrespondents in this second group were sent a printed version of the questionnaire. The first mailing for the third group offered the option of completing the questionnaire by either Web or by a printed version of the questionnaire. This group received a letter providing them the link to a location of a Web site with the questionnaire and a printed version of the questionnaire. We compared response rates at different stages of data collection, item nonresponse, and responses across the three modes. We also compare the 2009 results with a study of the same population conducted in 2007.