Ensuring quality in graduate teaching is a guarantee of the competitiveness demonstrated by institutions. The aim of this study was to determine the internal consistency and reliability of the Triple Jump (TJ) instrument applied to Master's and Doctorate students in Medical Sciences to assess compliance of goals. Reliability study conducted at Universidad de La Frontera, Temuco, Chile. TJ was applied by 3 evaluators who performed 2 evaluations, with structured assessment guidelines in three domains (cognitive, attitudinal and procedural), on a Likert scale with scores from 0 (not achieved) to 4 (achieved). The total point score was transformed into a final grade. The evaluated subjects were characterized, and internal consistency and interobserver reliability were determined. Fifty-one Health care professionals enrolled in Medical Sciences graduate programs at the Universidad de La Frontera between 2005-2016 (35 from the Master ’s program and 16 from the Doctorate in Medical Sciences). In study subjects had a median age of 29 years, 6 years of professional experience and 3 years as specialists. 66.7 % were male. The internal consistency of the responses to the assessment was characterized by a Cronbach’s between 0.734 and 0.938. Interobserver reliability of the instrument had values between 0.86 and 1.0. The levels of internal consistency and reliability were higher than previously reported; therefore, the test is reliable and exhibits solid internal consistency.