The Wisebag™, a lunchbag-style container with an electronic events-monitoring system, was designed as a real-time indirect objective measure of microbicide gel use. Due to cost, alternative functionalities (i.e. use of offline and dummy versions) were explored. We conducted a three-arm, double-blinded pilot study among 50 HIV-negative women in Durban, South Africa to assess participant adherence and Wisebag acceptability and performance. Participants were randomized 2:2:1 to Wisebag with online (events transmitted via cellular signal in real-time), offline (events stored in device memory) or inactive "dummy" devices. Participants were instructed to open the Wisebag daily for 2 weeks, retrieve a study sticker and affix it on a diary card. All participants completed the study. At exit, 94 % did not know which device they had received, nor could they differentiate the Wisebag types when presented with the three options. Five offline devices failed (no data recorded). Per Wisebag events, 26 % of women were perfectly adherent compared to 48 % by self-report and 46 % per diary card. Of reported non-adherence, 92 % did not open the Wisebag (travelling or forgot) and 22 % opened Wisebag >1×/day (curiosity). Participants liked and were comfortable carrying Wisebag. Successful blinding will allow inclusion of offline and/or dummy Wisebags in future study designs. Perfect adherence by opening events was significantly lower than by self-report, highlighting the importance of objective measures of adherence in clinical trials. Additional studies to validate Wisebag data with actual products, with and without SMS and online functionality, in different populations and settings, and in comparison to biomarkers are warranted.