In an effort to maintain or encourage higher response rates, an increasing number of surveys are including a Web option. Up until 2002, NSSATS was a multi-mode mail survey with telephone follow-up. A Web option was introduced in 2002, and to date, there have been three survey administrations with a Web option (2002, 2003, and 2004). In that time, the percentage of sample members who have chosen to respond by Web has increased nearly 75 percent, from around 17 percent to nearly 30 percent. Despite these gains in the percentage of Web respondents, how successful have these Web efforts really been overall? Using data from N-SSATS, we will look at such questions as: (1) Is including a Web option increasing response rates or are the Web respondents being plucked from other the existing modes; (2) To what extent are Web respondents enamored of this option, that is, once a Web respondent, do you tend to remain a Web respondent? and (3) how does the accuracy of the data provided by Web respondents compare to the accuracy provided by respondents in the other modes?