This report presents the best practices of Persistent Identifiers (PIDs) for end users, as defined by seven integrated use case partners participating in the FAIR-IMPACT project. These partners represent diverse scientific domains, including photon and neutron sciences, life sciences, and earth & environmental sciences. The use cases span different stages of the research life cycle, demonstrating the critical role of PIDs in processes such as data production workflows, complex data citation events, and the management of metadata and sensitive data. Key topics explored include versioning, granularity, kernel metadata, machine-actionability, and research reproducibility in relation to PID minting practices. Currently, there is a lack of standardised cross-domain guidance and widely accepted best practices on PID usage. This report seeks to address this gap by providing clear, adaptable, and applicable recommendations across multiple scientific domains.