A loss of previously established skills in the first three years of life, known as regression, may be an early feature of autism and thus a significant red flag in development. Yet, despite its diagnostic importance, regression stays a puzzling phenomenon, especially with regard to the nature of the skill losses, the association with other characteristics and developmental trajectories or long-term outcomes. Remarkably, there is growing phenotypical, neurobiological and genetic evidence for a link with stereotypic or repetitive motor movements. However, both research on regression and on motor stereotypies in autism is characterised by inconsistencies due to operationalisation issues and lack of long-term follow-up. The aim of the present project is to examine if early regression is associated with higher rates of motor stereotypies and atypical developmental trajectories resulting in more difficulties in later functioning. To obtain representative results, cohorts of children and adolescents with or at elevated likelihood for autism will be included in both cross-sectional and prospective longitudinal follow-up studies. In addition, different innovative technological devices will be implemented. More insight into the operationalisation of regression and the establishment of growth charts with outcomes related to both regression and motor stereotypies in autism, could impact aetiological knowledge as well as clinical guidelines on early detection and support.