The linear teleconnections between the ENSO phenomeno and the Cauca Valley-Colombia monthly interannual and seasonal hydrometeorology ((jan-feb-mar(EFM), apr-may-jun (AMJ), jul-aug-sep (JAS) and oct-nov-dec (OND)), was studied. Two multivaritate statistical techniques was used: empirical ortoghonal functions (EOFS) and the canonical correlation analysis (CCA). EOFS or
Principal Component analysis (PC) was applied to data monthly precipitation (1972-1998) of 50 raingauge, to data monthly flow (1951-2000) of 8 hydrometric stations of the Cauca Valley-Colombia, and to 12 macroclimatic variables asocciated to ENSO (EL NiŅO - Southern Oscilation) phenomeno. Previously, the raingauge stations were grouped in 3 homogeneous groups, applying an analysis of hierarchical cluster that was verified with the geographical method and the discriminant analysis of the first 4 EOFS of precipitation. They stand
out the advantages of the method EOFS to reduce the dimensionality of multivariate data to calculate gappy data, to evaluate and to reduce multicolineality, to conform homogeneous groups and to detect outliers.
To select the significant PC of each group of variables, 7 selection approaches were used: The graphic method, the percentage of explained variance, the root average, the Velicer test, Bartlett test, Broken Stick test and cross validation, being chosen this last as the best, because is more robust and more quantitative.
Five macroclimatic PC were selected, four of flow group, and for the homogenous precipitations groups, 1, 2 and 3: one, two and one, PC was selected, respectively, for a posterior canonical correlation analysis (CCA).

The ENSO effect is bigger in flows that precipitation. It was concluded that
in AMJ and OND periods the ENSO association and effect with the regional hydrometeorology is less; while in EFM and JAS, the effect is bigger.
When includes the principal components of macroclimatics variables like variable
predictors in the precipitation and flows models, improved in the prediction was obtained, indicating that they contribute with additional information. The flows models presented good adjustment, for what they can be used for previi
diction. Likewise, The multivariate EOFS and CCA methods proved to be a valuable tools in the study of climate variability to understand the relationships between the ENSO phenomeno with the region hydrometeorology.