Abstract The objective was to identify the internal structure of the measurement scale of the loneliness ESTE, created in Spain and applied in the population of older adults in the city of Saltillo, Coahuila, México. Loneliness has been conceptualized as perceived social isolation and has been described as a condition that can become serious, however, it has been seen as a subjective experience of lack of affection, intimacy, and closeness with family and/or friendships, as well as the difference between the preferred level and the actual level of social contact a person has. Lack of support networks and cohabitation in the homes of older adults generate conditions of vulnerability, which results in a weakness in the health issue, as well as being exposed to greater fragility that prevents maintaining the usual pace of life. The study was quantitative with instrumental design. We assessed the internal structure of the instrument through exploratory factor analysis (AFE) and confirmatory factorial analysis (AFC) in a convenience sample of 396 adults 55.8 % women and 44.2 % men from 60 to 95 years (M = 69.81, DE = 6.96) attending medical service in the State Center of the Elderly and in the programs of the Commission of Human Rights of the state of Coahuila, Mexico. The EAST scale measuring loneliness in older adults was used, with 34 items grouped into four factors family loneliness, marital loneliness, social loneliness, and existential crisis. The response format is ordinal with five options ranging from Never to Always. In addition, three questions were included regarding the perception of health. The McDonald's coefficient for instrument reliability was estimated for each of the dimensions. The analysis of the data was an Exploratory Factor Analysis (EFA) and the Confirmatory Factor Analysis (CFA). 79.8 % believe that their health is good health, this percentage being higher than those who consider their health to be bad and lousy. Compared to a person their age, 62.4 % of older adults are considered equal, 30.3 % feel better, and 6.8 % consider themselves worse, with differences in these three response groups. In conclusion, they were asked if they were satisfied with their lives, to which 94.4 % said yes. The four-factor solution explained 53.45 % of the variance and in the parallel analysis the total explained variance was 57.3 %. The factors that most variance explained are social loneliness (24.31 %) family solitude (12.13 %) and existential crisis (10.18 %). The marital loneliness factor was the least explained component (6.81 %). The ordinal alpha coefficient resulted with high reliability (ω = .920). Four factors were obtained with 22 indicators. According to the indicators, the model with correlated factors presented a close approximation between the hypothetical model and the data, this according to the value obtained in the RMSEA, which was .050, which is of good adjustment, the CFI presented a good fit since a value of .974 was reported, values between .950 and .990 are considered appropriate. The GFI and AGFI had values greater than .90, which represent good adjustment. Additionally, the performance of variables from sociodemographic groups showed no significant differences in totals estimates per factor. The validated scale for Mexican population is a useful tool to measure the level of loneliness. This study obtained evidence of the validity of the internal structure of the scale and estimated the reliability of the ESTE loneliness scale adapted to the Mexican population.