Automated Vehicles (AVs) are expected to bring many benefits to society (e.g. improved safety, increased capacity, reduced fuel use and emissions). However, there are also many challenges with AVs. These include issues associated with drivers’ trust, mental models of the automation’s capabilities and limitations and manual driving skill degradation. Therefore, solutions are needed to enhance the benefits and eliminate the challenges with AVs. One solution is driver training. Current training for drivers of AVs is limited to an owner’s manual and most drivers do not read their owner’s manual. Therefore, this thesis sought to understand the training needs for drivers of a Level 4 AV and to design, develop and evaluate a comprehensive training programme to address these needs. A grounded theory approach was used to identify nine key themes in AV driver training. These themes were applied to currently deployed training programmes, five AV collisions and IAM RoadSmart’s Advanced Driver Course to demonstrate the validity and relevance of these themes to AVs and driver training. A Training Needs Analysis (TNA) was conducted to establish the tasks and competencies that drivers need to safely operate the Level 4 AV. This TNA identified 7 main tasks, 25 sub-tasks, 2428 operations and 105 training needs and was used to develop an online video-based training resource and a training package for the safe activation of the Level 4 AV. Evaluation studies demonstrated short-term benefits of these training programmes over no training (more correct decisions, better activation behaviours) and owner’s manuals (more appropriate mental models, reduced mental demand), however the long-term retention benefits and applications to Level 5 AVs and other transport domains must be explored. This thesis should encourage further research into the development of better training for drivers of AVs, so that clear benefits of AVs can be realised without the challenges.