Aim. To develop a clinical core set of outcome measures that is accepted for relevance, feasibility and validity by stakeholders and useful for a) interaction between patient and professional, b) internal quality improvement, and c) external transparency in patients with NSLBP in primary care physical therapy. Method. We used a consensus-driven modified RAND-UCLA Delphi technique. We conducted seven separate steps with panellists (physical therapists, patient representatives, health insurers) to select accepted outcomes. These seven steps consisted of a literature search, two online surveys, patient interviews, an experts meeting, a consensus meeting and final approval of an advisory board. Results of previous steps were discussed during the consensus meeting, and then panellists voted for inclusion per measure. The final core set was rated on relevance and feasibility on a 9-point Likert scale, when the median was ≥7 the core set was accepted. Results. 34 panellists in two online surveys, five panellists in an expert committee, ten patients for semi-structured interviews and 26 panellists in a consensus meeting participated in the study. 12 outcome measures were rated and discussed and finally six outcome measures were accepted. The final core set was accepted with a median of 7. Conclusion. This study present an outcome set that is accepted by stakeholders as having added value for a) interaction between patient and professional, b) internal quality improvement, and c) external transparency in patients with NSLBP in primary care physical therapy. In a next project this outcome set will be tested on his reliability and feasibility in a large pilot. No conflicts of interest. Sources of funding: Health insurance company CZ, the Netherlands