OBJECTIVE: The role of religious involvement in mental health has been increasingly investigated in psychiatric research; however, there is a shortage of scales on religiousness in Portuguese. The present study aimed to develop and validate a brief instrument to assess intrinsic religiosity (Intrinsic Religiousness Inventory - IRI) in two Brazilian samples. METHOD: The initial version was based on literature review and experts' suggestions. University students (sample 1; n = 323) and psychiatric patients (sample 2; n = 102) completed the Duke Religiosity Index (DUREL), the IRI, an instrument of spirituality measurement (WHOQOL-SRPB), as well as measurements of anxiety and depressive symptoms. RESULTS: The IRI showed adequate internal consistence reliability in sample 1 (Cronbach's α = 0.96; 95% CI; 0.95-0.97) and sample 2 (α = 0.96; 95% CI; 0.95-0.97). The IRI main component analyses indicated a single factor, which explained 73.7% and 74.9% of variance in samples 1 and 2, respectively. Strong correlations between IRI and intrinsic subscale of the DUREL were observed (Spearman's r ranging from 0.87 to 0.73 in samples 1 and 2, respectively, p 0.70). CONCLUSION: These data indicate that the IRI is a valid instrument and may contribute to study intrinsic religiosity in Brazilian samples.