Dynamic Bayesian networks (DBNs) are powerful tools for modeling and studying complex dynamical systems. In the healthcare sector, they have been used for studying gene regulatory networks, to model disease progression and for making treatment decisions on patients suffering from chronic diseases. Despite their potential, the lack of comprehensive and publicly available software has strongly limited their use. In fact, to the best of our knowledge, no open-source software is publicly available for discrete dynamic Bayesian networks. In this paper we introduce and describe DBNcare, the first R package to address this gap. DBNcare provides an all-in-one framework for defining and visualizing network structures, performing parameter and structure learning, and enabling probabilistic inference and forecasting. A use case of a simplified diabetes management system, taken from the healthcare literature, is used to show how DBNcare allows to efficiently manage temporal dependencies and uncertainty. Source code, detailed documentation, and examples are made available via the DBNcare repository page.