The Burn-Specific Pain Anxiety Scale (BSPAS) estimates pain-related anxiety and determines the effect of treatment in patients with burns, especially regarding wound care. This study aimed to analyze the 9-item and the abbreviated 5-item BSPAS by the Rasch model. This prospective study included 161 patients admitted to Dutch burn centres. The BSPAS was administered during hospital stay resulting in 314 self-reports and was analysed using the Rasch unidimensional measurement model 2030 (RUMM 2030). Unidimensionality of the 9-item and 5-item BSPAS was confirmed. Initially, both versions did not fit the model due to response dependency. After creating subtests, fit to the model improved. After deleting 'feeling insecure about my healing' and creating two subtests with three items, fit of the 9-item BSPAS was obtained, while the 5-item BSPAS fitted after creating a subtest with two items. The Rasch model demonstrated that both versions were unidimensional and were able to fit the model after adjusting for response dependency. Moreover, the 5-item BSPAS could be further improved by deleting 'worrying about the possible pain.' A 4-item abbreviated BSPAS (BSPAS-4I) captures pain-related anxiety and is proposed to be used in future studies and daily practice.