Introduction The search for appropriate tools to assess communicational skills remains an ongoing challenge. The Calgary-Cambridge Observational Guide is designed for measuring and comparing performance in communication skills training Objective Adapt the 28-item version of the guide for the Brazilian cultural context and perform a psychometric quality analysis of the tool. Method Experienced preceptors (35) evaluated videos of 5 medical residents with a simulated patient. using the translated version. For the cultural adaptation we followed the methodological norms on synthesis, retro-translation, committee review, and testing. We conducted the data analysis with the Rasch Many-Facet Model. Results The reliability of the internal consistency was satisfactory, with the Cronbach's alpha coefficient acceptable in all 5 evaluations (0.88; 0.84; 0.89; 0.87; 0.83). The reliability coefficient was high in the evaluators’ facets (0.90), the stations (0.99), and items (0.98). The data indicates the suitability of the tool to the Rasch Many-Facet Model. The evaluators had greater difficulty with attitudinal items, such as demonstration of respect, confidence, and empathy. Conclusion The psychometric properties of the tool were adequate. The reliability indicators of the Rasch Model presented a good potential for reproducing the Brazilian version of the tool as well as acceptable reliability for its use.