High rates of Internet use among young adults make online intervention with this population particularly attractive. However, low adherence rates limit the exposure to and the potential effectiveness of these programs. This study identifies strategies for increasing adherence by examining the rates of participation for a 5-week beta (pilot) version and final version of the RealU Web site, an online intervention for college smokers. Three modifications from the beta to the RealU Web site were (a) changing format from a smoking cessation Web site to an online college life magazine, (b) providing proactive peer e-mail support, and (c) adopting a more linear site structure. Participants were recruited via Internet health screening and received US$10 for completing weekly study activities. Enrollment among eligible smokers was higher for the beta compared with the RealU intervention (47/69, 68.1% vs. 517/1618, 32.0%, p<.001), but participants did not differ in terms of age, gender, or past 30-day cigarette or alcohol use. Participation fell sharply during the beta test (53% in week 1 to 26% by week 5) compared with the RealU average of 95% (range 89% to 98%). Participation during each study's final week was much higher in the RealU (93% week 20) compared with the beta (26% week 5, p<.001). After 5 weeks, self-reported 30-day abstinence was higher for RealU intervention participants (16.0%) compared with the beta participants (4.3%, p5.03). The modifications from the beta to RealU Web site described above resulted in high rates of sustained participation over 20 weeks.