The analysis of social networks (SNA) constitutes a major advance in tourism research by revealing the characteristics of established relationships, displaying their structures and properties. This work aims to characterize and analyze the cooperation network formed between owners of inns in Tiradentes-MG. In order to explore the context of the survey, the interviews were conducted with seven owners of inns. Subsequently, a questionnaire was submitted to the owners for a data collection. A network of cooperation was created through the ‘snow ball’ technique. Going forward SNA techniques were used, with the emphasis on descriptive metrics and by generating Exponential Random Graph Models (ERGM). Thus it was possible to characterize the cooperation network and analyze its properties. The network has 54 inns, with a low overall density, because there are only 4.7% of possible relationships. 17 central and 37 peripheral establishments were identified. In the central subgroup the density was 15.4% and in the peripheral subgroup only 2.6%. The observed cooperation network presented homophily by gender and origin of the owners. The results of ERGM modeling allowed probabilistic explanations in terms of shared attributes of owners, such as gender and origin.