Reputation systems are fundamental for assessing the quality of user contributions inparticipatory sensing. However, naively associating reputation scores to contributionsallows adversaries to establish links between multiple contributions and thus deanonymizeusers. We present the IncogniSense framework as a panacea to these privacythreats. IncogniSense utilizes periodic pseudonyms generated using blind signatureand relies on reputation transfer between these pseudonyms. Simulations are used toanalyze various reputation cloaking schemes that address the inherent trade-off betweenanonymity protection and loss in reputation. Our threat analysis confirms the robustnessof IncogniSense and a prototype demonstrates that associated overheads are minimal.