The spatial-temporal variability characterizing the water quality of high Andean lagoons for tourist use was evaluated using multivariate statistical methods during 2017 and 2018. The water samples were collected from 14 sampling sites, with three replicates each. The water quality indicators determined were: pH, temperature, DO, COD, BOD5, P, N, Fe, Cu, Cr, Cd, Pb, Zn and chlorophyll-a. The flat cluster analysis (k R cluster) according to Ward's algorithm showed six significantly differentiated groups (α=0.01). In turn, the real similarity profile (SIMPROF) moves markedly away from the obtained low permutation with a large excess of Euclidean similarity with a Pi value of 0.627. The PCA showed that the first two components recommended by the sedimentation analysis (Scree test) indicated 61.52% of the total variation of the observations. According to the Spearman range correlation selection criterion, the variables that best interpret the sample distributions are COD, DTS, P, Cd and Zn with a correlation of 0.893, the DTS being the most important variable with a correlation value of 0.795. The PERMANOVA analysis according to the flat cluster factor indicated that at least one of the groups is different from the others in relation to the levels of physicochemical characteristics studied. Therefore, all the configured groups are statistically different, demonstrating that each lagoon is different in relation to its physicochemical indicators, according to the season in which it is found.