Fluctuation of groundwater level induces changes in pore-water pressure of soil. However, this variation is not considered for underground constructions. This article explores the application of a statistical method to evaluate the groundwater level variation in geotechnical designs. The methodology included: (i) data collection, (ii) statistic formulation, and (iii) statistic data analysis. We collected information from the technical studies of the project “Metro de Bogotá”, and selected four boreholes spanning 160 m, approximately, where the 1° de Mayo metro station will be built, in the south of the city. We used groundwater level readings reported by different piezometers for 30 days and data variance was assessed using a multivariate statistical method: analysis of repeated measures profiles. Results present a procedure to estimate the groundwater level fluctuation during a short monitoring period. We concluded that the analysis of repeated measures profiles allows estimating the groundwater level variation under a significance level 1-a.