The occurrence of genetic erosion in local maize varieties in Mexico is intensely debated. Recent publications from Mexico show contradicting results about the loss of local varieties. Genetic erosion is a complex process, and well-documented examples of actual genetic erosion are not common in the literature. We worked in a region in which adoption of improved varieties was negligible, but other factors affecting maize agriculture were at play. The objectives of the study were to describe changes in maize diversity in the last 10 years and to associate them with socio-economic and environmental changes in a region in Mexico’s Central Highlands. We used richness and abundance of local varieties and diversity indices of races as indicators of maize diversity changes over time. We analyzed statistics and based on interviews we evaluated maize diversity changes between 2005 and 2015. We interviewed 113 farmers on two occasions with intervals from 5 to 10 years. According to climate statistics, rain has declined and temperature has increased. We also found a decrease in the lake level during the past 35 years. The total population in the region has doubled since the 1960s. The indigenous population has not changed significantly. Number of people working in agriculture has decreased since the 1960s. Rain fed agriculture decreased 8.1 % from 1990 to 2007. In four villages studied, farmed land area had decreased between 1995 and 2015. This reduction varies between 22 and 39 % depending on the village. Maize planted area decreased from 9675 to 8115 ha from 2003 to 2014. In the same period, avocado plantations grew from 34 to 786 ha. In despite of these changes, we did not find significant changes in average landraces per farmer (2.13 ± 0.28 in 2015) nor per village (4.15 ± 1.26 in 2015). Significant changes in maize races were not found either (1.91 ± 0.26 per farmer, 2.85 ± 0.86 per village in 2015). These results show that maize landrace diversity in the region is resilient but dynamic.