This paper summarizes randomized experiment to study the effects of an Internetbased intervention on type 2 diabetes patients in Montevideo, Uruguay. The intervention consisted of a specially designed website and an electronic social network where participants were able to navigate freely, download materials, and interact with other diabetics and with specialists. No significant impact was found on participants` knowledge, behavior, or health outcomes. It was also found that only a minority of patients logged on to the website, and most were only reached by email and mobile text (SMS). Participation in the website is correlated with patients` characteristics, such as gender, marital status, and education.