High incidence of elderly depression motivates use of screening depression scales in aging; the most used scales by specialized and nonspecialized healthcare staff is Geriatric Depression Scale. Short version of this scale (15 items) was propose in 1982 whit a high effort in his construction and with no modification since there. The aim of this study was to determine the confiability of these 15 item-short version and ultrashort versions (5, 3, and 2 items). This study used information of the Encuesta Salud Bienestar y Envejecimiento SABE in Bogotá/Colombia (1957 registers). Statistical analysis showed a high variability in homogeneity items’ Cronbach analysis (between 0,59 until 0,23). Factorial analysis showed 3 factors in the scale. This seems to be part of a need to contextualize the scale because the use of it since 1982 without changes and a dynamic depression concept through the years.