Background: Working alliance between the physiotherapist and patient is important for successful treatment in physiotherapy. Working alliance can be measured by a 12-item self-reporting measure with 7-point Likert scale, the Working Alliance Inventory Short-Form Client (WAI-SC). Objectives: To develop a shorter version of the Japanese WAI-SC through examination of factor structure and test-retest reliability. Design: Survey. Method: A convenience sample of 118 outpatients with musculoskeletal disorders completed the WAI-SC immediately after the third consultation and 101 of them completed the WAI-SC immediately before the fourth consultation. Results/findings: Through investigations of internal consistency, test-retest reliability, and factor structure, a single structure was detected by excluding Item 4 and Item 10. Score calculation was also modified to dichotomous variables (Score 7 or other scores) due to serious ceiling effects, where >50% of responses were Score 7. A five-item questionnaire with dichotomous score rating (Modified WAI-SC Japanese musculoskeletal version) was developed through consideration of internal consistency and item-total correlation analysis. This questionnaire demonstrated adequate internal consistency (Cronbach's α = 0.88) and excellent test-retest reliability for each item (Cohen's κ for all > 0.4) and total scores (intraclass correlation coefficient = 0.84). Conclusions: The five-item Modified WAI-SC Japanese musculoskeletal version was developed with a single factor structure showing adequate internal consistency and excellent test-retest reliability.