While technology has the potential to improve treatment of offenders in forensic mental healthcare, it is not yet clear how, why and for whom it can be of added value in this unique context. This thesis aims to provide more insight into how eHealth interventions can be used in treatment of people who have committed aggressive or sexual offenses. Furthermore, attention is paid to how to optimally develop, implement and evaluate eHealth interventions with and for vulnerable and complex target groups. In this thesis’ first part, the current state of affairs of technology in treatment of offenders was analysed. A systematic review showed that most studies focused on language-based interventions, and that not much is known yet about technology in treatment of offenders. An analysis of the implementation of a web-based intervention in forensic outpatient care showed that it was not used as much as expected, which could be partly explained by a suboptimal fit between the characteristics of the technology, the skills of its users, and its integration in existing care structures. This thesis’ second part is focused on the participatory development of a virtual reality intervention. The iterative development process was guided by the CeHRes Roadmap and resulted in an interactive, personalized VR application to identify and treat triggers for delinquent behaviour. It became clear that VR can add something ‘new’ to treatment that cannot be achieved in in-person care, mostly by facilitating a focus on doing instead of talking. Finally, the third part of this thesis investigates the possibilities of another type of technology – a persuasive mobile app – to strengthen self-control, an important predictor of crime. This new self-control training app was evaluated by means of a mixed-methods approach in a student sample, and only self-control in the participants that used the app improved over time. The main conclusion of this thesis is that in order for eHealth to be of added value, adaptivity is key: the technology, the people involved, and the context in which it is used, but also the development, implementation and evaluation processes all have to be adaptive and flexible.