Since none of the existing systematic reviews of the literature related to smart cities have focused on smart cities applications using context-aware features, this study aims to provide a comprehensive view of (1) the characteristics of context-aware applications supported by smart cities’ infrastructures, (2) the characteristics of the context-aware technologies being used, and (3) the maturity level of the identified applications. An electronic search was conducted on Scopus, Web of Science, and IEEE Xplore databases, combining relevant keywords. After the studies’ selection, 27 studies were included. The results show the interest in using context-aware features to develop smart cities’ applications targeting public health, tourism experience, urban mobility, active citizenship, shopping experience, management of urban infrastructures, public alerts, recommenders, and smart environments. Moreover, the applications reported by the included studies do not allow the execution of actions without the explicit consent of the users, and they integrate personal sensors, sensors deployed in the city and crowdsourcing mechanisms to acquire primary context attributes (i.e., location, activity, and environment) and secondary context attributes (i.e., speed, physical activity, points of interest, stress, and tourists’ emotions). In addition, the included studies present scarce details about context reasoning, data privacy, integrity, and confidentiality, data aggregation and accuracy, and data interoperability. Finally, the results also indicate low maturity level of the developed applications, which is evidenced by the lack of user-centred evaluations, as well as by the difficulty in integrating important technologies being developed for smart cities.