AIMS: To develop an instrument for measuring thriving among residents in long-term care facilities, to assess the correspondence between proxy ratings and self-report and the internal consistency of the Norwegian version. BACKGROUND: The instrument was developed from the life-world concept of thriving and thereby has a different theoretical basis than existing 'dementia related' quality-of-life instruments. Thriving relates the experience of older persons to the place where they live. Proxy instruments need to be developed for residents in long-term care facilities who are not able to report their subjective experiences. DESIGN: Instrument development using cross-sectional survey design. METHODS: The instrument was developed in three versions (resident, family and staff) from a theory on thriving. Forty-eight triads consisting of a resident, family member and primary nurse from 12 Norwegian nursing homes participated. Data collection took place between March-December 2011. Inter-rater agreement between the groups was assessed by Cohen's kappa coefficient (weighted). Internal consistency was evaluated by Cronbach's alpha. Homogeneity was explored through item-total correlations. RESULTS: Agreement between residents, family members and staff was poor or fair (<0.41) in six of 38 items. These items were excluded. The 32-items instrument had satisfactory Cronbach's alpha values in each of the three samples and satisfactory homogeneity as item-total correlations was substantial without being excessive and thus indicated that items were measuring the same construct. CONCLUSION: The instrument appears to have internal consistency and enable reliable proxy measures of the thriving construct. Further psychometric assessment including checking for possible item redundancy is needed.