In several scientific and engineering disciplines, models have been used to understand the behavior of dynamic processes that evolve in space and in time by providing a probabilistic framework to analyzing the available information. The geostatistical tools used to analyze space-time data are based on established statistical methods, where time is considered as an additional dimension. These models have become very useful in fields such as meteorology, hydrology, ecology, geosciences, and environmental sciences, among others. Subsidence generated by the intense extraction of groundwater in a region is a dynamic phenomenon that manifests itself through the sinking of the ground surface, leading to significant settling in buildings and public utilities as well as cracks in roads. Since the regional subsidence of Mexico City is one of the most representative cases of this type in the world, in this work a model with a full grid space-time layout (STF) is used to analyze and predict the evolution of this phenomenon in the city, taking into account a monitoring system composed of 1931 surface benchmarks. Results show that the separable variogram model was the one that best represented the spatial and temporal correlation of the phenomenon in the area of study. In addition, the differences between the registered ground elevation made in 2016 and those estimated by the space-time model for the same year, were less than 1.00 m. This implies that in general accurate ground elevation values and subsidence rates can be obtained from the proposed space-time model during the time period 2010-2030 for the lacustrine zone of Mexico City.