Background: loneliness, a common feature among elderly people, can affect health and quality of life. Objective: this article examined affective bond and marital status among elders in "Barra do Garças", municipality, Mato Grosso State, Midwest Brazil. Methodology: from the 3,593 elderly people, 300 engaged in the study which had been approved by the ethical committee on research. Results: there were more illiterate men than women (37.7% and 21.6%, respectively). In respect of income source there was a relative contingent of women with no income comparing to the men. Considering the socioeconomic class the majority of elderly, 68% belong to the less fortunate socioeconomic classes (D and E). Conclusion: it has been observed a higher contingent of elderly people with no affective bond or living alone which influence the quality of life of those elderly by increasing the psychosocial as well as the health-related risks, which was widely discussed according to the world's scientific literature. city presents 9,141,84 Km 2 of area and is 510Km far from Cuiaba, the state's capital. The city population is 58,398 inhabitants and 3,593 people have 65 years old or more [2]. The studied population was constituted by people with 65 years old or more which attended at leisure-time activities offered by Social Assistance Secretary (SAS) of Barra do Garças municipality in partnership with the Federal University of Mato Grosso (UFMT). The convenience sample (n=300) included all subjects with 65 years old and more which engaged in leisure-time activities and agreed to be interviewed. Information regarding marital status, and socioeconomic variables such as gender, age, income and other information was collected. This transversal study was carried out during May and September 2015. People with 65 years old, engaged in the social activities, and that agree to engage into the study were included. The exclusion criteria was being less than 65 years old and those who did not agree to participate. This study did not present any kind of risk for the participants and the interview ensured the privacy of each subject. All participants signed the informed consent and the study was approved by the Ethics Committee on Research of the Julio Müller University Hospital (UFMT) (665/CEP-HUJM/09). Data was statistically analyzed by the EPI-INFO ® software (6.04d). The hypotheses of association between variables was accepted when the found p value was lower than 0.05. Methodology The current study was done in Barra do Garças municipality, Mato Grosso State, located at the border of Goias state. The Barra do Garças