This paper describes the process of revising the Belgian Nursing Minimum Data Set (NMDS). The study started in 2000. Implementation is planned from 2006. The project is divided in 4 major phases. The first phase (June – October 2002) implied the development of the conceptual framework based on literature review and secondary data-analysis. The Nursing Interventions Classification (NIC) was selected as framework for the revision of the NMDS. The second phase focused on the language development (November 2002 – September 2003) with panels of clinical experts (N=75) for six care programs. They indicated hospital financing, nurse staffing allocation and assessment of the appropriateness of hospitalization as priorities of a revised B-NMDS. A draft instrument with 84 variables, using NIC, was developed during this period. This leads to an alpha version of a revised NMDS. The third phase (October 2003 – December 2004) focused on the data collection and validation of the new tool. The new NMDS was tested on 158 nursing wards in 66 Belgian hospitals from December 2003 until March 2004. This test generated data for some 95.000 inpatient days. The interrater-reliability of the revised NMDS is tested. The criterion-related validity of the revised NMDS is compared with the actual NMDS. The discriminative power of the revised NMDS is tested to select the most relevant items for data collection. This will result in a beta version of revised NMDS in December 2004. The records of the revised NMDS are linked with the hospital discharge dataset and other mandatory datasets to integrate the revised NMDS in the broader health care management. The fourth phase (January – December 2005) will focus on information management.