Ensuring the representativeness of primary data sources is essential for evaluating cycling interventions, modelling cycling behaviour, and informing urban mobility planning. This study investigates the notion of representativeness within cycling-related data by analysing three primary sources: surveys, automatic counting stations, and GPS tracking data. A methodological framework is proposed to evaluate data quality, which is subsequently applied to datasets from the city of Dresden, Germany. The study demonstrates that no single data source can be considered universally superior; rather, the appropriateness of each source depends on the specific planning context. GPS-based cycling data proves particularly beneficial for network planning and traffic monitoring owing to its granularity and volume, while Household Travel Surveys (HTS)s remain indispensable for capturing socio-demographic variation, travel diaries, cycling mode share and at least in some cases subjective mobility experiences. The results underscore the importance of mixed-method approaches to mitigate the limitations inherent in individual datasets and to ensure comprehensive evidence bases for the planning of cycling infrastructure. Future research should aim to enhance the demographic inclusivity of digital data sources and to further refine validation methodologies to improve overall data reliability.