Objective: The Montgomery-Åsberg Depression Rating Scale (MADRS) is widely used to assess depression severity. The Structured Interview Guide for the MADRS (SIGMA) was created to standardize MADRS assessment. The objective of this study was to translate and validate the original SIGMA into a Brazilian Portuguese version (SIGMA-VB). Methods: We translated and cross-culturally validated the original SIGMA into the SIGMA-VB, and assessed its psychometric properties using data from 93 adult outpatients enrolled in the Integral Assessment in Unipolar Depression (AIUNI) trial. Participants were assessed by two raters on five visits over 8 weeks. We calculated multiple interrater reliability indexes for the SIGMA-VB and used the Hamilton Depression Hating Scale (HAM-D) for validation purposes. Results: According to the SIGMA-VB, participants had moderate depression at baseline followed by mild depression at 8 weeks. We found over 90% of correlation between scores attributed by different raters using the SIGMA-VB. Correlations between the SIGMA-VB and the HAM-D were above 66%. Conclusion: Our findings confirm that the SIGMA-VB is a valid and reliable instrument to assess depression severity in clinical research and practice. Its interrater reliability was similar to that of a previously published Japanese version of the SIGMA.