Using Chilean administrative datasets for the period 2011-2017, we study which of the most used tools to evaluate teacher quality, namely teachers' evaluation tests (TET) and teacher's value added (TVA), predicts more accurately not only short run (as most of the literature focus on) but also middle run students' outcomes. For this evaluation we follow the same cohorts of students and teachers. Our results suggest that the correlation between (TET) and (TVA) appears to be null in school outcomes. However, our analysis also reveals that both measures, TET and TVA, positively affect the probability of tertiary education attendance, indicating that both measures are complementary in measuring teacher quality in the middle run. These results have relevance from the public policy point of view as unlike countries (e.g. USA) where TVA is used for teacher's promotions and personnel decisions, in countries where TVA is not used for teacher's personnel decisions (e.g. Chile), TVA seems to be useful to measure teacher quality. Furthermore, our findings are consistent with the argument of the multidimensionality of teaching quality, because even though in the short run TVA and TET seem to be orthogonal, in the medium run they seem to be complementary tools to measure teacher effectiveness.