Based on the results from ICFES tests (Colombian equivalent of the American SATs) from 1980to 2009, the study uses a hierarchical/multi-level analysis to provide an estimate of the schooleffect in each educational sector for three different tests. The results suggest that the publicschool effect has always been smaller than that of private schools in the last thirty years andthat the effect on the mathematics section of the test is always smaller, regardless of the educationalsector. They also indicate that since 2000 there has been a considerable drop in thecoefficient, which could be related to the change in the tests that year. Lastly, an intertemporalanalysis of score percentiles reveals that private schools have traditionally ranked higher thantheir public counterparts.