Aims: To evaluate the psychometric properties of the Multidimensional Pain Inventory (MPI) in a Brazilian sample of patients with orofacial pain. Methods: A total of 1,925 adult patients, who sought dental care in the School of Dentistry of São Paulo State University's Araraquara campus, were invited to participate; 62.5% (n = 1,203) agreed to participate. Of these, 436 presented with orofacial pain and were included. The mean age was 39.9 (SD = 13.6) years and 74.5% were female. Confirmatory factor analysis was conducted using χ2/df, comparative fit index, goodness of fit index, and root mean square error of approximation as indices of goodness of fit. Convergent validity was estimated by the average variance extracted and composite reliability, and internal consistency by Cronbach's alpha standardized coefficient (α). The stability of the models was tested in independent samples (test and validation; dental pain and orofacial pain). The factorial invariance was estimated by multigroup analysis (Δχ2). Results: Factorial, convergent validity and internal consistency were adequate in all three parts of the MPI. To achieve this adequate fit for Part 1, item 15 needed to be deleted (λ = 0.13). Discriminant validity was compromised between the factors "activities outside the home" and "social activities" of Part 3 of the MPI in the total sample, validation sample, and in patients with dental pain and with orofacial pain. A strong invariance between different subsamples from the three parts of the MPI was detected. Conclusion: The MPI produced valid, reliable, and stable data for pain assessment among Brazilian patients with orofacial pain. Keywords: facial pain, pain measurement, psychometrics, validation studies