The first documented day-long program to encourage girls to consider computing as a career was held in 1987 in the U.K. Over the last 20 years these one-day events, labeled "Girls into Computing" days, have been conducted by academics and professionals to foster female-student interest in information technology (IT) degrees and careers. This paper charts the growth and popularity of these types of events internationally and questions their effectiveness. The format of "Girls in Computing" days varies from giving students hands-on experiences with technology to exposure to women in the industry through seminars and presentations. They are mostly aimed at girls in the 12-16 year old age bracket. The underlying assumption of these events is that female students are not choosing IT degrees and careers because of a limited awareness of what an IT degree and career involves. The absence of any longitudinal evaluation on the success of "Girls into Computing" events is a glaring oversight. Success of previous events conducted in the UK, USA, and Australia has been evaluated mainly through pre-event and post-event attitudinal type surveying of students. These have all been positive, but any long-term attitudinal change has not been measured and some researchers suggest that the effects of "Girls in Computing" days have been negligible, a suggestion supported by a continued decline in female IT higher education enrolment statistics in all these countries. In Australia there has been a plethora of these types of events, some catering to over 1000 girls a day, some smaller and more personalized. Information about these events is presented in tabular form in the paper. This is followed by a questioning of how to measure success or, indeed, if it can be measured at all. The researchers ask if the amount of time and effort, presented as costs and benefits, is worth it, given that there has been no change in the steady decline of women in IT. The paradox of how to measure the effectiveness of an event that attracts large international sponsorship and is presented to over 1000 girls a day over a smaller, localized, and personalized event with limited sponsorship is posed as a question that is yet to be answered. The sustainability of energy of the volunteers who run these events, women in IT, not the dominant gender of the discipline, is also presented as a paradox. This gender paradox complicates the evaluation of whether size matters in the construction of events to increase awareness around IT careers. A strategic evaluation of the measurement of success of "Girls in Computing" days needs to be adopted. There is no quick solution, and longitudinal tracking of student attendees and their future degree and career choices is complicated by privacy and funding issues. There is one example in the paper of an Australian follow-up proposal that is due next year. A survey of students who participated in a 2006 event is planned by the authors for 2009. It is concluded that we cannot afford to dismiss these days outright. We cannot conclude whether large events are better than smaller events until we have more robust data. Finally we must keep encouraging the dominant gender of the discipline to take a greater interest in the ownership of this issue for sake of the depth of diversity and creativity in the IT profession. (Contains 4 tables and 2 figures.)