As companies increasingly develop data products and follow data-driven strategies for competitive advantage, the decisions about sharing data become important. In practice, however, organizations are hesitant to participate in data sharing. To address this problem, this paper presents a taxonomy to characterize data assets and improve structured assessment of data assets for decisions on inter-organizational data sharing. The aim of the taxonomy is to aid organizations engaged in data spaces or data ecosystems in making informed decisions whether to share data assets and under what terms, based on their value, possible risks and the resulting criticality associated with sharing. To develop the taxonomy, we conducted a multivocal literature review, synthesizing scientific literature related to data sharing, descriptions of data assets used in real-world use cases, and shared in data space initiatives. The taxonomy structures data key dimensions and characteristics important for understanding business value and risks associated with data sharing. The taxonomy is an initial result of a method for data assessment, and lays groundwork for future theorizing on inter-organizational sharing of sensitive data assets.