At the bottom of the introduction page, the participants were asked to click the continue buttonif they had read through the information, were 18 or older, and consented to participate in the study (the formalconsent procedure).By clickingthe button,the participants were linked to the questionnaire hosted on Entergates server. The web questionnaire comprised a total of 85 questions. However, by using skip patterns, we were able to personalize the questions to some extent. This could reduce the number of questionsto 51 for some participants.Each participant opened a session on the server and replies were recorded on the server, withthe possibilityto reviseanswers (i.e.,the participants could go back and forth in the questionnaire before finally submitting them). Due to the format and technique used, we were able to identify those who dropped out before completion. A total of 4,637 participants commenced the questionna

Whenthe participants had finished or quitthe questionnaire,the session was shut down. Those who completed their questionnaire werelinkedto aweb pagethankingthem fortheir participation and encouragementtocontactthe researchersifthey hadany questions. None did so. Finally, they were linked to the research programs web site. Measure The measure was based on an earlier onethat we usedin a similar study in 2002 (see Cooper, Mansson, Daneback, Tikkanen, & Ross, 2003) but revised and expanded. The questionnaire comprised 85 questions divided into six subsections. The first section had 24 sociodemographic questions, including items on internet access, various forms of relationshipconstellations,love,and sexuality. Section two had 28 questions about love and sexual activities on the internet, focusing on what activities the participants engagedin, how frequently, and why. Sectionthree comprised 11 questions of sexual exposure

The fourth section had 6 questions about sexual activities in exchange for money or goods (attitudes and experiences). Section five had six questions about positive and negative experiences from using the internet forloveand sexual purposes.The sixth sectioncomprised 10 questions on online sexual problems (loss of control, addiction to online sex, and sexually transmitted infections and HIV). The five questions comprising the Internet sexual problems (ISP) measure appearinTable 2, and were scored on a four-pointLikert scale of problems (none, some, few, big). Before submission, the participants were provided with a possibility to comment on the questionnaire in an open ended question (text box). Analysis Data were analyzed using SPSS version 17. Analysis consisted of frequency distributions of demographics by gender and descriptive statistics for the ISP scale items. Cronbachs a was computed on the five items of the scale and the items formed into a scale with

Logistic regression (simultaneous entry) was carried out onthe binary scale score by demographic (gender, degree of religiosity) and sexual behavior (frequency of masturbation, had negative experiences using Internet forlove/sexual