The aim of this pilot study was to explore how students’ perceived risk of self-disclosing varies by different topics andrecipients. The current study adapted the 10 topics of the Self Disclosure Index (SDI; Miller, Berg & Archer, 1983) andmodified the way they were presented. The adapted SDI also extended the original SDI’s two recipients (a same-sex strangerand a same-sex friend) to a total of six (a close family member, a distant family member, a close friend, a distant friend,stranger/acquaintance, and a health professional). One hundred and eighty three American college students responded to anonline survey containing demographic questions, the adapted SDI and the original SDI. Data were analyzed using PASW/SPSS18. Statistical tests were run for internal consistency, construct validity, convergent validity, and known-group validity. It wasfound that the adapted SDI was internally consistent and reliable and had a high correlation with the original SDI. Four factorsemerged, and each factor was named Family, Strangers/Acquaintances, Health Professional, and Friends. According to theseresults, level of risk perceived depended more on who a person talked to rather than what a person talked about. Finally, resultsfrom MANOVA and the t-test showed a trend with risk and gender, which is consistent with previous research findings.Implication for future research for a development of a new scale of self-disclosure risks was discussed.