Training to enhance healthcare practitioners' capabilities in engaging people in sensitive and end-of life-related conversations is in demand. However, evaluations have either not measured, or found very limited impact on actual practice and patient experience. Training effectiveness is improved when it is based on in-depth evidence, reflects the complexity of real-life interactions, and instils principles adaptable to everyday practice. A relatively new source of in-depth evidence and practice-relevant insights on healthcare interactions is conversation analytic research, a form of observational analysis of real-life interactions. However, conversation analytic research findings have largely been disseminated by and for scientists, rather than clinicians and trainers. We used conversation analytic evidence to develop resources for use by healthcare trainers. The aim was to increase training's evidence-base and authenticity. We further aimed to develop resources applicable to working with learners ranging from novices to advanced practitioners. Methods: Using an intervention development approach, we created online video-clips and supplementary written materials for professionals who deliver training, supervision, and support in healthcare communication for staff and students. The materials were reviewed by an advisory group comprising clinicians, lay consultees, educators, and researchers, and piloted by trainers in UK universities, NHS organisations and independent hospices. We refined materials based on their feedback. Results: The resulting 'RealTalk' resources focus on practices for communicating with patients and their companions about end-of-life and prognosis. Two core training modules were developed, each comprising several patient case studies featuring video-clips from real-life healthcare consultations. The clips featured practices that patients and experienced practitioners use in approaching end-of-life matters. The case studies also included evidence-based descriptions of observable practices and the principles underlying these, alongside transcripts and case synopses. Conclusions: RealTalk training resources aim to facilitate evidence-based, experiential and reflective learning, focusing on communication challenges, practices and principles for end-of-life-related interactions. The resources are designed for use by trainers for delivering all levels of training, from introductory to advanced, in both formal and informal training settings. Our development process may serve as a blueprint for the production of future evidence-based training resources based on conversation analytic research.