Although there are a variety of objective instruments for assessing digital competencies, only a few are specifically designed to assess digital competencies that are relevant for the design and implementation of digitally supported natural science teaching. With a few exceptions, such as assessments based on complex classroom observations or portfolio analyses, research on these subject-specific digitalization-related competencies relies predominantly on self-assessment measures. However, self-assessment measures can be effectively embedded in the teaching context and serve as a competency framework that raises students' awareness of the digital skills they are expected to develop during their studies. Based on DiKoLAN CH, the Swiss adaptation of the DiKoLAN framework, a task-based test is currently being developed to diagnose subject-specific, digitization-related competencies for natural science teaching. The presentation provides an overview of the test format, reports on the results of expert validation, and discusses preliminary results from implementation in teacher training programs.